Instructions to use Scale-or-Reason/Qwen2.5-1.5B-math-ift with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Scale-or-Reason/Qwen2.5-1.5B-math-ift with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Scale-or-Reason/Qwen2.5-1.5B-math-ift")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Scale-or-Reason/Qwen2.5-1.5B-math-ift")
model = AutoModelForCausalLM.from_pretrained("Scale-or-Reason/Qwen2.5-1.5B-math-ift")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Scale-or-Reason/Qwen2.5-1.5B-math-ift with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Scale-or-Reason/Qwen2.5-1.5B-math-ift"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Scale-or-Reason/Qwen2.5-1.5B-math-ift",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Scale-or-Reason/Qwen2.5-1.5B-math-ift

SGLang

How to use Scale-or-Reason/Qwen2.5-1.5B-math-ift with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Scale-or-Reason/Qwen2.5-1.5B-math-ift" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Scale-or-Reason/Qwen2.5-1.5B-math-ift",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Scale-or-Reason/Qwen2.5-1.5B-math-ift" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Scale-or-Reason/Qwen2.5-1.5B-math-ift",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Scale-or-Reason/Qwen2.5-1.5B-math-ift with Docker Model Runner:
```
docker model run hf.co/Scale-or-Reason/Qwen2.5-1.5B-math-ift
```

Qwen2.5-1.5B-math-ift / slurm.out

Nicolas-BZRD

Add files using upload-large-folder tool

187fcc2 verified 10 months ago

Raw

History Blame Contribute Delete

216 kB

	3: W0902 18:31:19.695000 3352467 torch/distributed/run.py:792]
	3: W0902 18:31:19.695000 3352467 torch/distributed/run.py:792] *****************************************
	3: W0902 18:31:19.695000 3352467 torch/distributed/run.py:792] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
	3: W0902 18:31:19.695000 3352467 torch/distributed/run.py:792] *****************************************
	2: W0902 18:31:19.695000 2766555 torch/distributed/run.py:792]
	2: W0902 18:31:19.695000 2766555 torch/distributed/run.py:792] *****************************************
	2: W0902 18:31:19.695000 2766555 torch/distributed/run.py:792] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
	2: W0902 18:31:19.695000 2766555 torch/distributed/run.py:792] *****************************************
	0: W0902 18:31:19.696000 801759 torch/distributed/run.py:792]
	0: W0902 18:31:19.696000 801759 torch/distributed/run.py:792] *****************************************
	0: W0902 18:31:19.696000 801759 torch/distributed/run.py:792] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
	0: W0902 18:31:19.696000 801759 torch/distributed/run.py:792] *****************************************
	1: W0902 18:31:19.696000 3191159 torch/distributed/run.py:792]
	1: W0902 18:31:19.696000 3191159 torch/distributed/run.py:792] *****************************************
	1: W0902 18:31:19.696000 3191159 torch/distributed/run.py:792] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.
	1: W0902 18:31:19.696000 3191159 torch/distributed/run.py:792] *****************************************
	0: [2025-09-02 18:31:52,971] [INFO] [axolotl.utils.schemas.validation.check_eval_packing:119] [PID:801836] [RANK:0] explicitly setting `eval_sample_packing` to match `sample_packing`[39m
	2: [2025-09-02 18:31:52,971] [INFO] [axolotl.utils.schemas.validation.check_eval_packing:119] [PID:2766631] [RANK:0] explicitly setting `eval_sample_packing` to match `sample_packing`[39m
	1: [2025-09-02 18:31:52,971] [INFO] [axolotl.utils.schemas.validation.check_eval_packing:119] [PID:3191235] [RANK:0] explicitly setting `eval_sample_packing` to match `sample_packing`[39m
	0: [2025-09-02 18:31:52,972] [INFO] [axolotl.utils.schemas.validation.hint_sample_packing_padding:218] [PID:801836] [RANK:0] Setting `pad_to_sequence_len: true` to prevent memory leaks when sample_packing[39m
	2: [2025-09-02 18:31:52,972] [INFO] [axolotl.utils.schemas.validation.hint_sample_packing_padding:218] [PID:2766631] [RANK:0] Setting `pad_to_sequence_len: true` to prevent memory leaks when sample_packing[39m
	3: [2025-09-02 18:31:52,972] [INFO] [axolotl.utils.schemas.validation.check_eval_packing:119] [PID:3352543] [RANK:0] explicitly setting `eval_sample_packing` to match `sample_packing`[39m
	1: [2025-09-02 18:31:52,972] [INFO] [axolotl.utils.schemas.validation.hint_sample_packing_padding:218] [PID:3191235] [RANK:0] Setting `pad_to_sequence_len: true` to prevent memory leaks when sample_packing[39m
	3: [2025-09-02 18:31:52,972] [INFO] [axolotl.utils.schemas.validation.hint_sample_packing_padding:218] [PID:3352543] [RANK:0] Setting `pad_to_sequence_len: true` to prevent memory leaks when sample_packing[39m
	0: [2025-09-02 18:31:57,072] [INFO] [axolotl.cli.config.load_cfg:245] [PID:801836] [RANK:0] config:
	0: {
	0: "activation_offloading": false,
	0: "auto_resume_from_checkpoints": true,
	0: "axolotl_config_path": "/lustre/fswork/projects/rech/dgo/udv55np/train/tmp/1756826505612058881.yaml",
	0: "base_model": "/lustre/fswork/projects/rech/qwv/udv55np/Qwen/Qwen2.5-1.5B_ift",
	0: "base_model_config": "/lustre/fswork/projects/rech/qwv/udv55np/Qwen/Qwen2.5-1.5B_ift",
	0: "batch_size": 16,
	0: "bf16": true,
	0: "capabilities": {
	0: "bf16": true,
	0: "compute_capability": "sm_90",
	0: "fp8": false,
	0: "n_gpu": 16,
	0: "n_node": 1
	0: },
	0: "chat_template": "qwen_25",
	0: "context_parallel_size": 1,
	0: "dataloader_num_workers": 16,
	0: "dataloader_pin_memory": true,
	0: "dataloader_prefetch_factor": 256,
	0: "dataset_prepared_path": "/lustre/fsn1/projects/rech/dgo/udv55np/dataset_math/Qwen3-235B-A22B/0",
	0: "dataset_processes": 192,
	0: "datasets": [
	0: {
	0: "chat_template": "tokenizer_default",
	0: "field_messages": "conversations",
	0: "message_property_mappings": {
	0: "content": "content",
	0: "role": "role"
	0: },
	0: "path": "/lustre/fswork/projects/rech/qwv/udv55np/dataset/math/hf/no_thinking_text/generator/default-d32b2cae8ea7e541/0.0.0",
	0: "trust_remote_code": false,
	0: "type": "chat_template"
	0: }
	0: ],
	0: "ddp": true,
	0: "deepspeed": {
	0: "bf16": {
	0: "enabled": true
	0: },
	0: "gradient_accumulation_steps": "auto",
	0: "gradient_clipping": "auto",
	0: "train_batch_size": "auto",
	0: "train_micro_batch_size_per_gpu": "auto",
	0: "wall_clock_breakdown": false,
	0: "zero_optimization": {
	0: "contiguous_gradients": true,
	0: "overlap_comm": true,
	0: "reduce_bucket_size": "auto",
	0: "stage": 3,
	0: "stage3_gather_16bit_weights_on_model_save": true,
	0: "stage3_param_persistence_threshold": "auto",
	0: "stage3_prefetch_bucket_size": "auto",
	0: "sub_group_size": 0
	0: }
	0: },
	0: "device": "cuda:0",
	0: "device_map": {
	0: "": 0
	0: },
	0: "dion_rank_fraction": 1.0,
	0: "dion_rank_multiple_of": 1,
	0: "env_capabilities": {
	0: "torch_version": "2.6.0"
	0: },
	0: "eval_batch_size": 1,
	0: "eval_causal_lm_metrics": [
	0: "sacrebleu",
	0: "comet",
	0: "ter",
	0: "chrf"
	0: ],
	0: "eval_max_new_tokens": 128,
	0: "eval_sample_packing": true,
	0: "eval_table_size": 0,
	0: "evals_per_epoch": 0,
	0: "flash_attention": true,
	0: "fp16": false,
	0: "gradient_accumulation_steps": 1,
	0: "gradient_checkpointing": true,
	0: "gradient_checkpointing_kwargs": {
	0: "use_reentrant": true
	0: },
	0: "learning_rate": 7e-06,
	0: "lisa_layers_attribute": "model.layers",
	0: "load_best_model_at_end": false,
	0: "load_in_4bit": false,
	0: "load_in_8bit": false,
	0: "local_rank": 0,
	0: "logging_steps": 10,
	0: "lora_dropout": 0.0,
	0: "loraplus_lr_embedding": 1e-06,
	0: "lr_scheduler": "warmup_stable_decay",
	0: "lr_scheduler_kwargs": {
	0: "min_lr_ratio": 0.1,
	0: "num_decay_steps": 300
	0: },
	0: "max_prompt_len": 512,
	0: "mean_resizing_embeddings": false,
	0: "micro_batch_size": 1,
	0: "model_config_type": "qwen2",
	0: "num_epochs": 1.0,
	0: "optimizer": "adamw_torch_fused",
	0: "output_dir": "/lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0",
	0: "pad_to_sequence_len": true,
	0: "pretrain_multipack_attn": true,
	0: "pretrain_multipack_buffer_size": 10000,
	0: "profiler_steps_start": 0,
	0: "qlora_sharded_model_loading": false,
	0: "ray_num_workers": 1,
	0: "resources_per_worker": {
	0: "GPU": 1
	0: },
	0: "sample_packing": true,
	0: "sample_packing_bin_size": 200,
	0: "sample_packing_group_size": 100000,
	0: "save_only_model": false,
	0: "save_safetensors": true,
	0: "save_steps": 0.2,
	0: "save_total_limit": 20,
	0: "sequence_len": 16384,
	0: "shuffle_before_merging_datasets": false,
	0: "shuffle_merged_datasets": true,
	0: "skip_prepare_dataset": false,
	0: "special_tokens": {
	0: "bos_token": "<\|im_start\|>",
	0: "eos_token": "<\|im_end\|>",
	0: "pad_token": "<\|endoftext\|>"
	0: },
	0: "strict": false,
	0: "tensor_parallel_size": 1,
	0: "tf32": false,
	0: "tiled_mlp_use_original_mlp": true,
	0: "tokenizer_config": "/lustre/fswork/projects/rech/qwv/udv55np/Qwen/Qwen2.5-1.5B_ift",
	0: "torch_dtype": "torch.bfloat16",
	0: "train_on_inputs": false,
	0: "trl": {
	0: "log_completions": false,
	0: "mask_truncated_completions": false,
	0: "ref_model_mixup_alpha": 0.9,
	0: "ref_model_sync_steps": 64,
	0: "scale_rewards": true,
	0: "sync_ref_model": false,
	0: "use_vllm": false,
	0: "vllm_server_host": "0.0.0.0",
	0: "vllm_server_port": 8000
	0: },
	0: "use_ray": false,
	0: "use_tensorboard": true,
	0: "val_set_size": 0.0,
	0: "vllm": {
	0: "device": "auto",
	0: "dtype": "auto",
	0: "gpu_memory_utilization": 0.9,
	0: "host": "0.0.0.0",
	0: "port": 8000
	0: },
	0: "warmup_steps": 150,
	0: "weight_decay": 0.0,
	0: "world_size": 16
	0: }[39m
	0: [2025-09-02 18:31:57,074] [INFO] [axolotl.cli.checks.check_user_token:35] [PID:801836] [RANK:0] Skipping HuggingFace token verification because HF_HUB_OFFLINE is set to True. Only local files will be used.[39m
	3: [2025-09-02 18:31:57,658] [INFO] [axolotl.utils.data.sft._load_raw_datasets:314] [PID:3352543] [RANK:0] Loading raw datasets...[39m
	3: [2025-09-02 18:31:57,926] [INFO] [axolotl.utils.data.wrappers.get_dataset_wrapper:88] [PID:3352543] [RANK:0] Loading dataset: /lustre/fswork/projects/rech/qwv/udv55np/dataset/math/hf/no_thinking_text/generator/default-d32b2cae8ea7e541/0.0.0 with base_type: chat_template and prompt_style: None[39m
	3: Tokenizing Prompts (num_proc=192): 0%\| \| 0/321773 [00:00<?, ? examples/s] Tokenizing Prompts (num_proc=192): 0%\| \| 1000/321773 [00:07<41:25, 129.06 examples/s] Tokenizing Prompts (num_proc=192): 1%\| \| 2000/321773 [00:08<18:32, 287.45 examples/s] Tokenizing Prompts (num_proc=192): 1%\| \| 3000/321773 [00:08<10:32, 504.34 examples/s] Tokenizing Prompts (num_proc=192): 2%\|▏ \| 5000/321773 [00:08<04:51, 1084.88 examples/s] Tokenizing Prompts (num_proc=192): 2%\|▏ \| 7000/321773 [00:08<03:01, 1731.38 examples/s] Tokenizing Prompts (num_proc=192): 2%\|▏ \| 8000/321773 [00:09<02:33, 2047.72 examples/s] Tokenizing Prompts (num_proc=192): 3%\|▎ \| 10000/321773 [00:09<01:40, 3116.78 examples/s] Tokenizing Prompts (num_proc=192): 3%\|▎ \| 11000/321773 [00:09<01:37, 3171.82 examples/s] Tokenizing Prompts (num_proc=192): 4%\|▍ \| 14000/321773 [00:09<00:56, 5450.93 examples/s] Tokenizing Prompts (num_proc=192): 5%\|▌
	3: \| 17000/321773 [00:09<00:39, 7709.86 examples/s] Tokenizing Prompts (num_proc=192): 6%\|▌ \| 19000/321773 [00:10<00:42, 7127.15 examples/s] Tokenizing Prompts (num_proc=192): 7%\|▋ \| 22000/321773 [00:10<00:30, 9898.27 examples/s] Tokenizing Prompts (num_proc=192): 7%\|▋ \| 24000/321773 [00:10<00:26, 11183.95 examples/s] Tokenizing Prompts (num_proc=192): 8%\|▊ \| 26000/321773 [00:10<00:24, 12209.62 examples/s] Tokenizing Prompts (num_proc=192): 9%\|▊ \| 28000/321773 [00:11<00:35, 8209.11 examples/s] Tokenizing Prompts (num_proc=192): 9%\|▉ \| 30000/321773 [00:11<00:42, 6838.15 examples/s] Tokenizing Prompts (num_proc=192): 11%\|█ \| 34000/321773 [00:11<00:27, 10590.54 examples/s] Tokenizing Prompts (num_proc=192): 11%\|█▏ \| 37000/321773 [00:11<00:21, 13142.37 examples/s] Tokenizing Prompts (num_proc=192): 12%\|█▏ \| 39000/321773 [00:11<00:20, 14089.58 examples/s] Tokenizing Prompts (num_proc=192): 13%\|█▎
	3: \| 41000/321773 [00:11<00:19, 14656.19 examples/s] Tokenizing Prompts (num_proc=192): 13%\|█▎ \| 43000/321773 [00:12<00:17, 15518.82 examples/s] Tokenizing Prompts (num_proc=192): 14%\|█▍ \| 45676/321773 [00:12<00:39, 6929.08 examples/s] Tokenizing Prompts (num_proc=192): 15%\|█▍ \| 47676/321773 [00:13<00:34, 7888.30 examples/s] Tokenizing Prompts (num_proc=192): 15%\|█▌ \| 49676/321773 [00:13<00:31, 8763.53 examples/s] Tokenizing Prompts (num_proc=192): 16%\|█▋ \| 52676/321773 [00:13<00:29, 9083.16 examples/s] Tokenizing Prompts (num_proc=192): 17%\|█▋ \| 54676/321773 [00:13<00:27, 9796.75 examples/s] Tokenizing Prompts (num_proc=192): 18%\|█▊ \| 56352/321773 [00:13<00:26, 9896.88 examples/s] Tokenizing Prompts (num_proc=192): 18%\|█▊ \| 58352/321773 [00:13<00:24, 10627.36 examples/s] Tokenizing Prompts (num_proc=192): 19%\|█▊ \| 59704/321773 [00:14<00:26, 10006.86 examples/s] Tokenizing Prompts (num_proc=192): 1
	3: 9%\|█▉ \| 61704/321773 [00:14<00:30, 8478.52 examples/s] Tokenizing Prompts (num_proc=192): 20%\|██ \| 64380/321773 [00:14<00:24, 10349.69 examples/s] Tokenizing Prompts (num_proc=192): 21%\|██ \| 66732/321773 [00:14<00:22, 11441.90 examples/s] Tokenizing Prompts (num_proc=192): 21%\|██▏ \| 68408/321773 [00:14<00:22, 11133.35 examples/s] Tokenizing Prompts (num_proc=192): 22%\|██▏ \| 70084/321773 [00:15<00:22, 11054.40 examples/s] Tokenizing Prompts (num_proc=192): 22%\|██▏ \| 72084/321773 [00:15<00:21, 11433.73 examples/s] Tokenizing Prompts (num_proc=192): 23%\|██▎ \| 73760/321773 [00:15<00:28, 8671.94 examples/s] Tokenizing Prompts (num_proc=192): 23%\|██▎ \| 75112/321773 [00:15<00:35, 6965.05 examples/s] Tokenizing Prompts (num_proc=192): 24%\|██▍ \| 77788/321773 [00:16<00:26, 9088.82 examples/s] Tokenizing Prompts (num_proc=192): 25%\|██▌ \| 80464/321773 [00:16<00:22, 10968.34 examples/s] Tokenizin
	3: g Prompts (num_proc=192): 27%\|██▋ \| 85816/321773 [00:16<00:14, 16672.80 examples/s] Tokenizing Prompts (num_proc=192): 28%\|██▊ \| 88492/321773 [00:16<00:17, 13055.51 examples/s] Tokenizing Prompts (num_proc=192): 28%\|██▊ \| 90844/321773 [00:16<00:17, 13538.06 examples/s] Tokenizing Prompts (num_proc=192): 29%\|██▉ \| 92520/321773 [00:16<00:17, 12750.13 examples/s] Tokenizing Prompts (num_proc=192): 29%\|██▉ \| 94872/321773 [00:17<00:16, 13495.26 examples/s] Tokenizing Prompts (num_proc=192): 31%\|███ \| 98900/321773 [00:17<00:12, 18153.54 examples/s] Tokenizing Prompts (num_proc=192): 32%\|███▏ \| 101900/321773 [00:17<00:10, 20426.88 examples/s] Tokenizing Prompts (num_proc=192): 32%\|███▏ \| 104576/321773 [00:17<00:10, 21175.21 examples/s] Tokenizing Prompts (num_proc=192): 33%\|███▎ \| 106928/321773 [00:17<00:10, 19557.61 examples/s] Tokenizing Prompts (num_proc=192): 34%\|███▍ \| 109280/321
	3: 773 [00:17<00:11, 18027.02 examples/s] Tokenizing Prompts (num_proc=192): 35%\|███▍ \| 111280/321773 [00:18<00:20, 10169.94 examples/s] Tokenizing Prompts (num_proc=192): 35%\|███▌ \| 113308/321773 [00:18<00:24, 8659.59 examples/s] Tokenizing Prompts (num_proc=192): 36%\|███▌ \| 114660/321773 [00:18<00:28, 7336.72 examples/s] Tokenizing Prompts (num_proc=192): 36%\|███▌ \| 116012/321773 [00:19<00:27, 7480.22 examples/s] Tokenizing Prompts (num_proc=192): 37%\|███▋ \| 117688/321773 [00:19<00:24, 8295.67 examples/s] Tokenizing Prompts (num_proc=192): 37%\|███▋ \| 119716/321773 [00:19<00:21, 9466.16 examples/s] Tokenizing Prompts (num_proc=192): 38%\|███▊ \| 121392/321773 [00:19<00:29, 6781.76 examples/s] Tokenizing Prompts (num_proc=192): 39%\|███▊ \| 124420/321773 [00:19<00:20, 9798.48 examples/s] Tokenizing Prompts (num_proc=192): 40%\|███▉ \| 128096/321773 [00:19<00:13, 13992.17 examples/s] Tokenizing
	3: Prompts (num_proc=192): 41%\|████ \| 130448/321773 [00:20<00:13, 14306.37 examples/s] Tokenizing Prompts (num_proc=192): 42%\|████▏ \| 134124/321773 [00:20<00:14, 13155.96 examples/s] Tokenizing Prompts (num_proc=192): 42%\|████▏ \| 135800/321773 [00:20<00:18, 10140.66 examples/s] Tokenizing Prompts (num_proc=192): 43%\|████▎ \| 139476/321773 [00:20<00:14, 12956.31 examples/s] Tokenizing Prompts (num_proc=192): 44%\|████▍ \| 141152/321773 [00:21<00:14, 12452.86 examples/s] Tokenizing Prompts (num_proc=192): 45%\|████▍ \| 143504/321773 [00:21<00:17, 10482.67 examples/s] Tokenizing Prompts (num_proc=192): 45%\|████▌ \| 145180/321773 [00:21<00:17, 10380.40 examples/s] Tokenizing Prompts (num_proc=192): 46%\|████▌ \| 147180/321773 [00:21<00:15, 11028.80 examples/s] Tokenizing Prompts (num_proc=192): 46%\|████▌ \| 148532/321773 [00:21<00:16, 10261.03 examples/s] Tokenizing Prompts (num_proc=192): 4
	3: 7%\|████▋ \| 152560/321773 [00:22<00:11, 14278.03 examples/s] Tokenizing Prompts (num_proc=192): 48%\|████▊ \| 154236/321773 [00:22<00:12, 13315.22 examples/s] Tokenizing Prompts (num_proc=192): 48%\|████▊ \| 155912/321773 [00:22<00:13, 12684.46 examples/s] Tokenizing Prompts (num_proc=192): 49%\|████▉ \| 157588/321773 [00:22<00:13, 12104.25 examples/s] Tokenizing Prompts (num_proc=192): 49%\|████▉ \| 159264/321773 [00:22<00:13, 11670.35 examples/s] Tokenizing Prompts (num_proc=192): 50%\|█████ \| 160940/321773 [00:22<00:14, 11283.84 examples/s] Tokenizing Prompts (num_proc=192): 51%\|█████ \| 163616/321773 [00:22<00:12, 12964.34 examples/s] Tokenizing Prompts (num_proc=192): 52%\|█████▏ \| 167968/321773 [00:23<00:08, 17147.06 examples/s] Tokenizing Prompts (num_proc=192): 53%\|█████▎ \| 171996/321773 [00:23<00:07, 20097.64 examples/s] Tokenizing Prompts (num_proc=192): 54%\|█████▍
	3: \| 174348/321773 [00:23<00:08, 16916.63 examples/s] Tokenizing Prompts (num_proc=192): 55%\|█████▌ \| 177024/321773 [00:23<00:09, 15704.11 examples/s] Tokenizing Prompts (num_proc=192): 56%\|█████▌ \| 178700/321773 [00:23<00:10, 14114.51 examples/s] Tokenizing Prompts (num_proc=192): 56%\|█████▌ \| 180376/321773 [00:24<00:10, 13461.92 examples/s] Tokenizing Prompts (num_proc=192): 57%\|█████▋ \| 182052/321773 [00:24<00:12, 10749.16 examples/s] Tokenizing Prompts (num_proc=192): 57%\|█████▋ \| 184052/321773 [00:24<00:12, 11135.40 examples/s] Tokenizing Prompts (num_proc=192): 58%\|█████▊ \| 186052/321773 [00:24<00:12, 10583.35 examples/s] Tokenizing Prompts (num_proc=192): 59%\|█████▊ \| 188404/321773 [00:24<00:11, 11790.53 examples/s] Tokenizing Prompts (num_proc=192): 59%\|█████▉ \| 190756/321773 [00:24<00:09, 13102.67 examples/s] Tokenizing Prompts (num_proc=192): 60%\|█████▉ \| 1
	3: 92432/321773 [00:25<00:09, 13422.19 examples/s] Tokenizing Prompts (num_proc=192): 61%\|██████ \| 195460/321773 [00:25<00:10, 11613.09 examples/s] Tokenizing Prompts (num_proc=192): 61%\|██████▏ \| 197812/321773 [00:25<00:09, 13512.01 examples/s] Tokenizing Prompts (num_proc=192): 62%\|██████▏ \| 199488/321773 [00:25<00:08, 13593.92 examples/s] Tokenizing Prompts (num_proc=192): 63%\|██████▎ \| 201840/321773 [00:25<00:08, 13750.64 examples/s] Tokenizing Prompts (num_proc=192): 64%\|██████▍ \| 206544/321773 [00:26<00:07, 14651.83 examples/s] Tokenizing Prompts (num_proc=192): 65%\|██████▍ \| 208220/321773 [00:26<00:09, 12434.73 examples/s] Tokenizing Prompts (num_proc=192): 65%\|██████▌ \| 209896/321773 [00:26<00:08, 12693.20 examples/s] Tokenizing Prompts (num_proc=192): 66%\|██████▌ \| 211572/321773 [00:26<00:08, 12889.40 examples/s] Tokenizing Prompts (num_proc=192): 67%\|██████
	3: ▋ \| 214248/321773 [00:26<00:07, 14194.53 examples/s] Tokenizing Prompts (num_proc=192): 67%\|██████▋ \| 216276/321773 [00:26<00:07, 15061.91 examples/s] Tokenizing Prompts (num_proc=192): 68%\|██████▊ \| 218628/321773 [00:27<00:08, 12262.17 examples/s] Tokenizing Prompts (num_proc=192): 68%\|██████▊ \| 220304/321773 [00:27<00:09, 10696.70 examples/s] Tokenizing Prompts (num_proc=192): 69%\|██████▉ \| 222304/321773 [00:27<00:08, 11855.66 examples/s] Tokenizing Prompts (num_proc=192): 71%\|███████ \| 227008/321773 [00:27<00:05, 18143.36 examples/s] Tokenizing Prompts (num_proc=192): 71%\|███████▏ \| 229360/321773 [00:27<00:06, 14199.06 examples/s] Tokenizing Prompts (num_proc=192): 73%\|███████▎ \| 235712/321773 [00:27<00:03, 21816.01 examples/s] Tokenizing Prompts (num_proc=192): 74%\|███████▍ \| 239388/321773 [00:28<00:03, 22456.78 examples/s] Tokenizing Prompts (num_proc=192): 76%\|�
	3: ��██████▌ \| 243064/321773 [00:28<00:03, 23999.96 examples/s] Tokenizing Prompts (num_proc=192): 76%\|███████▋ \| 246092/321773 [00:28<00:03, 21010.93 examples/s] Tokenizing Prompts (num_proc=192): 77%\|███████▋ \| 248444/321773 [00:28<00:04, 14670.92 examples/s] Tokenizing Prompts (num_proc=192): 78%\|███████▊ \| 251120/321773 [00:29<00:06, 10872.04 examples/s] Tokenizing Prompts (num_proc=192): 79%\|███████▊ \| 253120/321773 [00:29<00:06, 9999.79 examples/s] Tokenizing Prompts (num_proc=192): 80%\|███████▉ \| 256472/321773 [00:29<00:05, 12446.17 examples/s] Tokenizing Prompts (num_proc=192): 80%\|████████ \| 258148/321773 [00:29<00:06, 9825.03 examples/s] Tokenizing Prompts (num_proc=192): 81%\|████████▏ \| 261500/321773 [00:29<00:04, 12948.51 examples/s] Tokenizing Prompts (num_proc=192): 82%\|████████▏ \| 263528/321773 [00:30<00:04, 13505.24 examples/s] Tokenizi
	3: ng Prompts (num_proc=192): 83%\|████████▎ \| 265556/321773 [00:30<00:04, 12073.28 examples/s] Tokenizing Prompts (num_proc=192): 83%\|████████▎ \| 268232/321773 [00:30<00:03, 14660.58 examples/s] Tokenizing Prompts (num_proc=192): 84%\|████████▍ \| 270908/321773 [00:30<00:03, 13163.71 examples/s] Tokenizing Prompts (num_proc=192): 85%\|████████▌ \| 273936/321773 [00:31<00:04, 11479.23 examples/s] Tokenizing Prompts (num_proc=192): 87%\|████████▋ \| 279964/321773 [00:31<00:02, 18840.31 examples/s] Tokenizing Prompts (num_proc=192): 88%\|████████▊ \| 282964/321773 [00:31<00:01, 20146.33 examples/s] Tokenizing Prompts (num_proc=192): 89%\|████████▉ \| 286316/321773 [00:31<00:01, 19035.77 examples/s] Tokenizing Prompts (num_proc=192): 90%\|████████▉ \| 288668/321773 [00:31<00:02, 16189.10 examples/s] Tokenizing Prompts (num_proc=192): 90%\|█████████ \| 290696/321
	3: 773 [00:32<00:02, 11163.90 examples/s] Tokenizing Prompts (num_proc=192): 91%\|█████████ \| 292724/321773 [00:32<00:02, 11719.21 examples/s] Tokenizing Prompts (num_proc=192): 92%\|█████████▏\| 294752/321773 [00:32<00:03, 8280.15 examples/s] Tokenizing Prompts (num_proc=192): 92%\|█████████▏\| 296104/321773 [00:32<00:02, 8643.73 examples/s] Tokenizing Prompts (num_proc=192): 93%\|█████████▎\| 298132/321773 [00:32<00:02, 9182.47 examples/s] Tokenizing Prompts (num_proc=192): 93%\|█████████▎\| 299484/321773 [00:33<00:02, 9735.08 examples/s] Tokenizing Prompts (num_proc=192): 93%\|█████████▎\| 300836/321773 [00:33<00:02, 9589.84 examples/s] Tokenizing Prompts (num_proc=192): 94%\|█████████▍\| 302188/321773 [00:33<00:02, 8482.04 examples/s] Tokenizing Prompts (num_proc=192): 94%\|█████████▍\| 303540/321773 [00:33<00:02, 8937.65 examples/s] Tokenizing Prompts (num_p
	3: roc=192): 95%\|█████████▍\| 304892/321773 [00:33<00:02, 7545.40 examples/s] Tokenizing Prompts (num_proc=192): 95%\|█████████▌\| 306243/321773 [00:33<00:02, 7636.62 examples/s] Tokenizing Prompts (num_proc=192): 96%\|█████████▌\| 307593/321773 [00:34<00:01, 7924.16 examples/s] Tokenizing Prompts (num_proc=192): 96%\|█████████▌\| 309620/321773 [00:34<00:01, 7315.62 examples/s] Tokenizing Prompts (num_proc=192): 97%\|█████████▋\| 311646/321773 [00:34<00:01, 5983.22 examples/s] Tokenizing Prompts (num_proc=192): 97%\|█████████▋\| 312997/321773 [00:35<00:01, 5833.55 examples/s] Tokenizing Prompts (num_proc=192): 98%\|█████████▊\| 315698/321773 [00:35<00:00, 7810.36 examples/s] Tokenizing Prompts (num_proc=192): 99%\|█████████▊\| 317048/321773 [00:35<00:00, 7275.24 examples/s] Tokenizing Prompts (num_proc=192): 99%\|█████████▉\| 318398/321773 [00
	3: :35<00:00, 7960.74 examples/s] Tokenizing Prompts (num_proc=192): 99%\|█████████▉\| 319748/321773 [00:36<00:00, 6454.31 examples/s] Tokenizing Prompts (num_proc=192): 100%\|█████████▉\| 321098/321773 [00:36<00:00, 6991.26 examples/s] Tokenizing Prompts (num_proc=192): 100%\|██████████\| 321773/321773 [00:36<00:00, 8790.55 examples/s]
	3: Dropping Long Sequences (>16384) (num_proc=192): 0%\| \| 0/321773 [00:00<?, ? examples/s] Dropping Long Sequences (>16384) (num_proc=192): 0%\| \| 1000/321773 [00:01<09:09, 583.81 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 4%\|▍ \| 13000/321773 [00:01<00:31, 9748.54 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 8%\|▊ \| 24676/321773 [00:02<00:16, 18505.00 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 10%\|▉ \| 31436/321773 [00:02<00:16, 17386.79 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 11%\|█▏ \| 36492/321773 [00:02<00:15, 18238.78 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 13%\|█▎ \| 40844/321773 [00:02<00:13, 20542.41 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 14%\|█▍ \| 45196/321773 [00:02<00:11, 23529.08 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 15%\|█▌ \| 49224/321773 [00:03<00:10, 25848.89 exampl
	3: es/s] Dropping Long Sequences (>16384) (num_proc=192): 17%\|█▋ \| 53224/321773 [00:03<00:09, 27406.33 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 50%\|█████ \| 160968/321773 [00:03<00:00, 242231.63 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 64%\|██████▍ \| 205600/321773 [00:03<00:00, 283879.55 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 76%\|███████▌ \| 243444/321773 [00:03<00:00, 242398.75 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 86%\|████████▌ \| 275640/321773 [00:03<00:00, 227367.38 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 95%\|█████████▍\| 304213/321773 [00:03<00:00, 181908.04 examples/s] Dropping Long Sequences (>16384) (num_proc=192): 100%\|██████████\| 321773/321773 [00:04<00:00, 69680.11 examples/s]
	3: Drop Samples with Zero Trainable Tokens (num_proc=192): 0%\| \| 0/315947 [00:00<?, ? examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 0%\| \| 1000/315947 [00:01<09:07, 574.76 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 3%\|▎ \| 9000/315947 [00:01<00:46, 6579.69 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 5%\|▌ \| 17000/315947 [00:01<00:21, 13663.54 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 7%\|▋ \| 23292/315947 [00:02<00:15, 19157.30 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 9%\|▉ \| 29230/315947 [00:02<00:13, 21824.77 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 11%\|█ \| 34814/315947 [00:02<00:11, 24900.39 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 13%\|█▎ \| 40044/315947 [00:02<00:11, 24024.45 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 15%\|
	3: █▍ \| 45982/315947 [00:02<00:09, 29154.70 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 16%\|█▌ \| 51274/315947 [00:02<00:08, 31939.93 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 18%\|█▊ \| 56212/315947 [00:03<00:08, 30464.92 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 19%\|█▉ \| 60796/315947 [00:03<00:08, 29937.43 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 21%\|██ \| 66088/315947 [00:03<00:07, 33742.94 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 23%\|██▎ \| 71380/315947 [00:03<00:06, 37325.43 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 24%\|██▍ \| 76318/315947 [00:03<00:06, 35668.58 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 26%\|██▌ \| 81256/315947 [00:03<00:09, 23980.92 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 27%\|██▋ \|
	3: 84840/315947 [00:04<00:10, 23099.84 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 28%\|██▊ \| 89132/315947 [00:04<00:08, 25831.01 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 30%\|███ \| 95716/315947 [00:04<00:06, 33324.19 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 84%\|████████▍ \| 265082/315947 [00:04<00:00, 394437.69 examples/s] Drop Samples with Zero Trainable Tokens (num_proc=192): 100%\|██████████\| 315947/315947 [00:05<00:00, 57226.91 examples/s]
	3: Add position_id column (Sample Packing) (num_proc=192): 0%\| \| 0/315947 [00:00<?, ? examples/s] Add position_id column (Sample Packing) (num_proc=192): 0%\| \| 1000/315947 [00:02<10:54, 481.25 examples/s] Add position_id column (Sample Packing) (num_proc=192): 4%\|▍ \| 13000/315947 [00:02<00:37, 8147.50 examples/s] Add position_id column (Sample Packing) (num_proc=192): 7%\|▋ \| 23000/315947 [00:02<00:18, 15678.25 examples/s] Add position_id column (Sample Packing) (num_proc=192): 10%\|█ \| 32000/315947 [00:02<00:12, 23384.21 examples/s] Add position_id column (Sample Packing) (num_proc=192): 14%\|█▍ \| 44000/315947 [00:02<00:07, 35667.24 examples/s] Add position_id column (Sample Packing) (num_proc=192): 17%\|█▋ \| 55000/315947 [00:02<00:05, 46194.47 examples/s] Add position_id column (Sample Packing) (num_proc=192): 20%\|██ \| 64292/315947 [00:02<00:05, 49525.45 examples/s] Add position_id column (Sample Packing) (num_proc=192):
	3: 25%\|██▍ \| 77876/315947 [00:02<00:03, 63396.95 examples/s] Add position_id column (Sample Packing) (num_proc=192): 29%\|██▊ \| 90460/315947 [00:03<00:03, 73841.86 examples/s] Add position_id column (Sample Packing) (num_proc=192): 32%\|███▏ \| 100690/315947 [00:03<00:03, 71582.31 examples/s] Add position_id column (Sample Packing) (num_proc=192): 35%\|███▌ \| 111504/315947 [00:03<00:02, 77234.71 examples/s] Add position_id column (Sample Packing) (num_proc=192): 39%\|███▉ \| 123964/315947 [00:03<00:02, 87113.03 examples/s] Add position_id column (Sample Packing) (num_proc=192): 43%\|████▎ \| 134424/315947 [00:03<00:02, 84181.32 examples/s] Add position_id column (Sample Packing) (num_proc=192): 48%\|████▊ \| 152238/315947 [00:03<00:01, 107063.91 examples/s] Add position_id column (Sample Packing) (num_proc=192): 53%\|█████▎ \| 168114/315947 [00:03<00:01, 119038.24 examples/s] Add position_id column (Sample Packi
	3: ng) (num_proc=192): 57%\|█████▋ \| 181574/315947 [00:03<00:01, 121647.10 examples/s] Add position_id column (Sample Packing) (num_proc=192): 63%\|██████▎ \| 198326/315947 [00:03<00:00, 133616.17 examples/s] Add position_id column (Sample Packing) (num_proc=192): 67%\|██████▋ \| 212847/315947 [00:04<00:00, 117655.05 examples/s] Add position_id column (Sample Packing) (num_proc=192): 73%\|███████▎ \| 230660/315947 [00:04<00:00, 130519.18 examples/s] Add position_id column (Sample Packing) (num_proc=192): 78%\|███████▊ \| 247115/315947 [00:04<00:00, 139062.35 examples/s] Add position_id column (Sample Packing) (num_proc=192): 83%\|████████▎ \| 262212/315947 [00:04<00:00, 135531.58 examples/s] Add position_id column (Sample Packing) (num_proc=192): 88%\|████████▊ \| 276470/315947 [00:04<00:00, 131531.90 examples/s] Add position_id column (Sample Packing) (num_proc=192): 92%\|█████████▏
	3: \| 290147/315947 [00:04<00:00, 119964.83 examples/s] Add position_id column (Sample Packing) (num_proc=192): 96%\|█████████▌\| 303047/315947 [00:04<00:00, 108030.56 examples/s] Add position_id column (Sample Packing) (num_proc=192): 100%\|█████████▉\| 314657/315947 [00:05<00:00, 73525.53 examples/s] Add position_id column (Sample Packing) (num_proc=192): 100%\|██████████\| 315947/315947 [00:05<00:00, 57407.99 examples/s]
	3: Saving the dataset (0/192 shards): 0%\| \| 0/315947 [00:00<?, ? examples/s] Saving the dataset (0/192 shards): 1%\| \| 1646/315947 [00:01<05:32, 944.80 examples/s] Saving the dataset (1/192 shards): 1%\| \| 1646/315947 [00:01<05:32, 944.80 examples/s] Saving the dataset (2/192 shards): 1%\| \| 3292/315947 [00:01<05:30, 944.80 examples/s] Saving the dataset (3/192 shards): 2%\|▏ \| 4938/315947 [00:01<05:29, 944.80 examples/s] Saving the dataset (4/192 shards): 2%\|▏ \| 6584/315947 [00:01<05:27, 944.80 examples/s] Saving the dataset (5/192 shards): 3%\|▎ \| 8230/315947 [00:01<05:25, 944.80 examples/s] Saving the dataset (6/192 shards): 3%\|▎ \| 9876/315947 [00:01<05:23, 944.80 examples/s] Saving the dataset (7/192 shards): 4%\|▎ \| 11522/315947 [00:01<05:22, 944.80 examples/s] Saving the dataset (8/192 shards): 4%\|▍ \| 13168/315947 [00:01<05:20, 944.80 examples/s] Saving the dataset (9/192 shards): 5%\|▍
	3: \| 14814/315947 [00:01<05:18, 944.80 examples/s] Saving the dataset (10/192 shards): 5%\|▌ \| 16460/315947 [00:01<05:16, 944.80 examples/s] Saving the dataset (11/192 shards): 6%\|▋ \| 19752/315947 [00:01<05:13, 944.80 examples/s] Saving the dataset (12/192 shards): 6%\|▋ \| 19752/315947 [00:01<05:13, 944.80 examples/s] Saving the dataset (13/192 shards): 7%\|▋ \| 21398/315947 [00:01<05:11, 944.80 examples/s] Saving the dataset (14/192 shards): 7%\|▋ \| 23044/315947 [00:01<05:10, 944.80 examples/s] Saving the dataset (15/192 shards): 8%\|▊ \| 24690/315947 [00:01<05:08, 944.80 examples/s] Saving the dataset (16/192 shards): 8%\|▊ \| 26336/315947 [00:01<05:06, 944.80 examples/s] Saving the dataset (17/192 shards): 9%\|▉ \| 27982/315947 [00:01<05:04, 944.80 examples/s] Saving the dataset (18/192 shards): 9%\|▉ \| 29628/315947 [00:01<05:03, 944.80 examples/s] Saving the dataset (19/192 shards): 10%\|█ \| 32920/31594
	3: 7 [00:01<04:59, 944.80 examples/s] Saving the dataset (20/192 shards): 10%\|█ \| 32920/315947 [00:01<04:59, 944.80 examples/s] Saving the dataset (21/192 shards): 11%\|█▏ \| 36212/315947 [00:01<04:56, 944.80 examples/s] Saving the dataset (22/192 shards): 11%\|█▏ \| 36212/315947 [00:01<04:56, 944.80 examples/s] Saving the dataset (23/192 shards): 12%\|█▏ \| 37858/315947 [00:01<04:54, 944.80 examples/s] Saving the dataset (24/192 shards): 13%\|█▎ \| 41150/315947 [00:01<04:50, 944.80 examples/s] Saving the dataset (25/192 shards): 13%\|█▎ \| 41150/315947 [00:01<04:50, 944.80 examples/s] Saving the dataset (26/192 shards): 14%\|█▎ \| 42796/315947 [00:01<04:49, 944.80 examples/s] Saving the dataset (27/192 shards): 14%\|█▍ \| 44442/315947 [00:01<04:47, 944.80 examples/s] Saving the dataset (28/192 shards): 15%\|█▍ \| 46088/315947 [00:01<04:45, 944.80 examples/s] Saving the dataset (29/192 shards): 15%\|█▌ \| 47734/3
	3: 15947 [00:01<04:43, 944.80 examples/s] Saving the dataset (30/192 shards): 16%\|█▌ \| 51026/315947 [00:01<04:40, 944.80 examples/s] Saving the dataset (31/192 shards): 16%\|█▌ \| 51026/315947 [00:01<04:40, 944.80 examples/s] Saving the dataset (32/192 shards): 17%\|█▋ \| 52672/315947 [00:01<04:38, 944.80 examples/s] Saving the dataset (33/192 shards): 17%\|█▋ \| 54318/315947 [00:01<04:36, 944.80 examples/s] Saving the dataset (34/192 shards): 18%\|█▊ \| 55964/315947 [00:01<04:35, 944.80 examples/s] Saving the dataset (35/192 shards): 18%\|█▊ \| 57610/315947 [00:01<04:33, 944.80 examples/s] Saving the dataset (36/192 shards): 19%\|█▉ \| 59256/315947 [00:01<04:31, 944.80 examples/s] Saving the dataset (37/192 shards): 20%\|█▉ \| 62548/315947 [00:01<04:28, 944.80 examples/s] Saving the dataset (38/192 shards): 20%\|█▉ \| 62548/315947 [00:01<04:28, 944.80 examples/s] Saving the dataset (39/192 shards): 21%\|██ \| 6
	3: 5840/315947 [00:01<04:24, 944.80 examples/s] Saving the dataset (40/192 shards): 21%\|██▏ \| 67486/315947 [00:01<04:22, 944.80 examples/s] Saving the dataset (41/192 shards): 21%\|██▏ \| 67486/315947 [00:01<04:22, 944.80 examples/s] Saving the dataset (42/192 shards): 22%\|██▏ \| 69132/315947 [00:01<04:21, 944.80 examples/s] Saving the dataset (43/192 shards): 22%\|██▏ \| 70778/315947 [00:01<04:19, 944.80 examples/s] Saving the dataset (44/192 shards): 23%\|██▎ \| 72424/315947 [00:01<04:17, 944.80 examples/s] Saving the dataset (45/192 shards): 23%\|██▎ \| 74070/315947 [00:01<04:16, 944.80 examples/s] Saving the dataset (46/192 shards): 24%\|██▍ \| 75716/315947 [00:01<04:14, 944.80 examples/s] Saving the dataset (47/192 shards): 24%\|██▍ \| 77362/315947 [00:01<04:12, 944.80 examples/s] Saving the dataset (48/192 shards): 25%\|██▌ \| 79008/315947 [00:01<04:10, 944.80 examples/s] Saving the dataset (49/192 shards)
	3: : 26%\|██▌ \| 80654/315947 [00:01<04:09, 944.80 examples/s] Saving the dataset (50/192 shards): 27%\|██▋ \| 83946/315947 [00:01<04:05, 944.80 examples/s] Saving the dataset (51/192 shards): 27%\|██▋ \| 83946/315947 [00:01<04:05, 944.80 examples/s] Saving the dataset (52/192 shards): 27%\|██▋ \| 85592/315947 [00:01<04:03, 944.80 examples/s] Saving the dataset (53/192 shards): 28%\|██▊ \| 87238/315947 [00:01<04:02, 944.80 examples/s] Saving the dataset (54/192 shards): 28%\|██▊ \| 88884/315947 [00:01<04:00, 944.80 examples/s] Saving the dataset (55/192 shards): 29%\|██▊ \| 90530/315947 [00:01<03:58, 944.80 examples/s] Saving the dataset (56/192 shards): 29%\|██▉ \| 92176/315947 [00:01<03:56, 944.80 examples/s] Saving the dataset (57/192 shards): 30%\|██▉ \| 93822/315947 [00:01<03:55, 944.80 examples/s] Saving the dataset (58/192 shards): 30%\|███ \| 95468/315947 [00:01<03:53, 944.80 examples/s] Saving t
	3: he dataset (59/192 shards): 31%\|███ \| 97114/315947 [00:01<03:51, 944.80 examples/s] Saving the dataset (60/192 shards): 31%\|███▏ \| 98760/315947 [00:01<03:49, 944.80 examples/s] Saving the dataset (61/192 shards): 32%\|███▏ \| 100406/315947 [00:01<03:48, 944.80 examples/s] Saving the dataset (62/192 shards): 33%\|███▎ \| 103698/315947 [00:01<03:44, 944.80 examples/s] Saving the dataset (63/192 shards): 33%\|███▎ \| 103698/315947 [00:01<03:44, 944.80 examples/s] Saving the dataset (64/192 shards): 33%\|███▎ \| 105344/315947 [00:01<03:42, 944.80 examples/s] Saving the dataset (65/192 shards): 34%\|███▍ \| 106990/315947 [00:01<03:41, 944.80 examples/s] Saving the dataset (66/192 shards): 35%\|███▍ \| 110282/315947 [00:01<03:37, 944.80 examples/s] Saving the dataset (67/192 shards): 35%\|███▍ \| 110282/315947 [00:01<03:37, 944.80 examples/s] Saving the dataset (68/192 shards): 35%\|███▌ \| 111
	3: 928/315947 [00:01<03:35, 944.80 examples/s] Saving the dataset (69/192 shards): 36%\|███▋ \| 115220/315947 [00:01<03:32, 944.80 examples/s] Saving the dataset (70/192 shards): 36%\|███▋ \| 115220/315947 [00:01<03:32, 944.80 examples/s] Saving the dataset (71/192 shards): 37%\|███▋ \| 116866/315947 [00:01<03:30, 944.80 examples/s] Saving the dataset (72/192 shards): 38%\|███▊ \| 118512/315947 [00:01<03:28, 944.80 examples/s] Saving the dataset (73/192 shards): 39%\|███▊ \| 121804/315947 [00:01<03:25, 944.80 examples/s] Saving the dataset (74/192 shards): 39%\|███▊ \| 121804/315947 [00:01<03:25, 944.80 examples/s] Saving the dataset (75/192 shards): 39%\|███▉ \| 123450/315947 [00:01<03:23, 944.80 examples/s] Saving the dataset (76/192 shards): 40%\|███▉ \| 125096/315947 [00:01<03:22, 944.80 examples/s] Saving the dataset (77/192 shards): 40%\|████ \| 126742/315947 [00:01<03:20, 944.80 examples/s] Saving t
	3: he dataset (78/192 shards): 41%\|████ \| 128388/315947 [00:01<03:18, 944.80 examples/s] Saving the dataset (79/192 shards): 41%\|████ \| 130034/315947 [00:01<03:16, 944.80 examples/s] Saving the dataset (80/192 shards): 42%\|████▏ \| 131680/315947 [00:01<03:15, 944.80 examples/s] Saving the dataset (81/192 shards): 42%\|████▏ \| 133326/315947 [00:01<03:13, 944.80 examples/s] Saving the dataset (82/192 shards): 43%\|████▎ \| 134972/315947 [00:01<03:11, 944.80 examples/s] Saving the dataset (83/192 shards): 43%\|████▎ \| 136618/315947 [00:01<03:09, 944.80 examples/s] Saving the dataset (84/192 shards): 44%\|████▍ \| 138264/315947 [00:01<03:08, 944.80 examples/s] Saving the dataset (85/192 shards): 45%\|████▍ \| 141556/315947 [00:01<03:04, 944.80 examples/s] Saving the dataset (86/192 shards): 45%\|████▍ \| 141556/315947 [00:01<03:04, 944.80 examples/s] Saving the dataset (87/192 shards): 45%\|█�
	3: �██▌ \| 143202/315947 [00:01<03:02, 944.80 examples/s] Saving the dataset (88/192 shards): 46%\|████▋ \| 146494/315947 [00:01<02:59, 944.80 examples/s] Saving the dataset (89/192 shards): 46%\|████▋ \| 146494/315947 [00:01<02:59, 944.80 examples/s] Saving the dataset (90/192 shards): 47%\|████▋ \| 148140/315947 [00:01<02:57, 944.80 examples/s] Saving the dataset (91/192 shards): 47%\|████▋ \| 149786/315947 [00:01<02:55, 944.80 examples/s] Saving the dataset (92/192 shards): 48%\|████▊ \| 151432/315947 [00:01<02:54, 944.80 examples/s] Saving the dataset (93/192 shards): 48%\|████▊ \| 153078/315947 [00:01<02:52, 944.80 examples/s] Saving the dataset (94/192 shards): 49%\|████▉ \| 154724/315947 [00:01<02:50, 944.80 examples/s] Saving the dataset (95/192 shards): 49%\|████▉ \| 156370/315947 [00:01<02:48, 944.80 examples/s] Saving the dataset (96/192 shards): 50%\|█████ \| 158016/315947 [00
	3: :01<02:47, 944.80 examples/s] Saving the dataset (97/192 shards): 51%\|█████ \| 159662/315947 [00:01<02:45, 944.80 examples/s] Saving the dataset (98/192 shards): 52%\|█████▏ \| 162954/315947 [00:01<02:41, 944.80 examples/s] Saving the dataset (99/192 shards): 52%\|█████▏ \| 162954/315947 [00:01<02:41, 944.80 examples/s] Saving the dataset (100/192 shards): 52%\|█████▏ \| 164600/315947 [00:01<02:40, 944.80 examples/s] Saving the dataset (101/192 shards): 53%\|█████▎ \| 167892/315947 [00:01<02:36, 944.80 examples/s] Saving the dataset (102/192 shards): 53%\|█████▎ \| 167892/315947 [00:01<02:36, 944.80 examples/s] Saving the dataset (103/192 shards): 54%\|█████▎ \| 169538/315947 [00:01<02:34, 944.80 examples/s] Saving the dataset (104/192 shards): 54%\|█████▍ \| 171184/315947 [00:01<02:33, 944.80 examples/s] Saving the dataset (105/192 shards): 55%\|█████▌ \| 174476/315947 [00:01<02:29, 9
	3: 44.80 examples/s] Saving the dataset (106/192 shards): 56%\|█████▌ \| 176121/315947 [00:01<02:27, 944.80 examples/s] Saving the dataset (107/192 shards): 56%\|█████▌ \| 176121/315947 [00:01<02:27, 944.80 examples/s] Saving the dataset (108/192 shards): 56%\|█████▋ \| 177767/315947 [00:01<02:26, 944.80 examples/s] Saving the dataset (109/192 shards): 57%\|█████▋ \| 179412/315947 [00:01<02:24, 944.80 examples/s] Saving the dataset (110/192 shards): 57%\|█████▋ \| 181057/315947 [00:01<02:22, 944.80 examples/s] Saving the dataset (111/192 shards): 58%\|█████▊ \| 182702/315947 [00:01<02:21, 944.80 examples/s] Saving the dataset (112/192 shards): 58%\|█████▊ \| 184347/315947 [00:01<02:19, 944.80 examples/s] Saving the dataset (113/192 shards): 59%\|█████▉ \| 185992/315947 [00:01<02:17, 944.80 examples/s] Saving the dataset (114/192 shards): 59%\|█████▉ \| 187637/315947 [00:01<02:15, 944.80 e
	3: xamples/s] Saving the dataset (115/192 shards): 60%\|█████▉ \| 189282/315947 [00:01<02:14, 944.80 examples/s] Saving the dataset (116/192 shards): 60%\|██████ \| 190927/315947 [00:01<02:12, 944.80 examples/s] Saving the dataset (117/192 shards): 61%\|██████ \| 192572/315947 [00:01<02:10, 944.80 examples/s] Saving the dataset (118/192 shards): 61%\|██████▏ \| 194217/315947 [00:01<02:08, 944.80 examples/s] Saving the dataset (119/192 shards): 62%\|██████▏ \| 195862/315947 [00:01<02:07, 944.80 examples/s] Saving the dataset (120/192 shards): 63%\|██████▎ \| 197507/315947 [00:01<02:05, 944.80 examples/s] Saving the dataset (121/192 shards): 63%\|██████▎ \| 199152/315947 [00:01<02:03, 944.80 examples/s] Saving the dataset (122/192 shards): 64%\|██████▎ \| 200797/315947 [00:01<02:01, 944.80 examples/s] Saving the dataset (123/192 shards): 64%\|██████▍ \| 202442/315947 [00:01<02:00, 944
	3: .80 examples/s] Saving the dataset (124/192 shards): 65%\|██████▍ \| 204087/315947 [00:01<01:58, 944.80 examples/s] Saving the dataset (125/192 shards): 65%\|██████▌ \| 205732/315947 [00:01<01:56, 944.80 examples/s] Saving the dataset (126/192 shards): 66%\|██████▌ \| 207377/315947 [00:01<01:54, 944.80 examples/s] Saving the dataset (127/192 shards): 66%\|██████▌ \| 209022/315947 [00:01<01:53, 944.80 examples/s] Saving the dataset (128/192 shards): 67%\|██████▋ \| 210667/315947 [00:01<01:51, 944.80 examples/s] Saving the dataset (129/192 shards): 67%\|██████▋ \| 212312/315947 [00:01<01:49, 944.80 examples/s] Saving the dataset (130/192 shards): 68%\|██████▊ \| 215602/315947 [00:01<01:46, 944.80 examples/s] Saving the dataset (131/192 shards): 68%\|██████▊ \| 215602/315947 [00:01<01:46, 944.80 examples/s] Saving the dataset (132/192 shards): 69%\|██████▉ \| 218892/315947 [00:01
	3: <01:42, 944.80 examples/s] Saving the dataset (133/192 shards): 70%\|██████▉ \| 220537/315947 [00:01<01:40, 944.80 examples/s] Saving the dataset (134/192 shards): 70%\|██████▉ \| 220537/315947 [00:01<01:40, 944.80 examples/s] Saving the dataset (135/192 shards): 71%\|███████ \| 223827/315947 [00:01<01:37, 944.80 examples/s] Saving the dataset (136/192 shards): 71%\|███████▏ \| 225472/315947 [00:01<01:35, 944.80 examples/s] Saving the dataset (137/192 shards): 71%\|███████▏ \| 225472/315947 [00:01<01:35, 944.80 examples/s] Saving the dataset (138/192 shards): 72%\|███████▏ \| 227117/315947 [00:01<01:34, 944.80 examples/s] Saving the dataset (139/192 shards): 72%\|███████▏ \| 228762/315947 [00:01<01:32, 944.80 examples/s] Saving the dataset (140/192 shards): 73%\|███████▎ \| 230407/315947 [00:01<01:30, 944.80 examples/s] Saving the dataset (141/192 shards): 73%\|███████▎
	3: \| 232052/315947 [00:01<01:28, 944.80 examples/s] Saving the dataset (142/192 shards): 74%\|███████▍ \| 233697/315947 [00:01<01:27, 944.80 examples/s] Saving the dataset (143/192 shards): 74%\|███████▍ \| 235342/315947 [00:01<01:25, 944.80 examples/s] Saving the dataset (144/192 shards): 75%\|███████▌ \| 236987/315947 [00:01<01:23, 944.80 examples/s] Saving the dataset (145/192 shards): 76%\|███████▌ \| 238632/315947 [00:01<01:21, 944.80 examples/s] Saving the dataset (146/192 shards): 76%\|███████▌ \| 240277/315947 [00:01<01:20, 944.80 examples/s] Saving the dataset (147/192 shards): 77%\|███████▋ \| 241922/315947 [00:01<01:18, 944.80 examples/s] Saving the dataset (148/192 shards): 78%\|███████▊ \| 245212/315947 [00:01<01:14, 944.80 examples/s] Saving the dataset (149/192 shards): 78%\|███████▊ \| 245212/315947 [00:01<01:14, 944.80 examples/s] Saving the dataset (150/192 shards):
	3: 79%\|███████▊ \| 248502/315947 [00:01<01:11, 944.80 examples/s] Saving the dataset (151/192 shards): 79%\|███████▊ \| 248502/315947 [00:01<01:11, 944.80 examples/s] Saving the dataset (152/192 shards): 79%\|███████▉ \| 250147/315947 [00:01<01:09, 944.80 examples/s] Saving the dataset (153/192 shards): 80%\|████████ \| 253437/315947 [00:01<01:06, 944.80 examples/s] Saving the dataset (154/192 shards): 80%\|████████ \| 253437/315947 [00:01<01:06, 944.80 examples/s] Saving the dataset (155/192 shards): 81%\|████████ \| 255082/315947 [00:01<01:04, 944.80 examples/s] Saving the dataset (156/192 shards): 81%\|████████▏ \| 256727/315947 [00:01<01:02, 944.80 examples/s] Saving the dataset (157/192 shards): 82%\|████████▏ \| 260017/315947 [00:01<00:59, 944.80 examples/s] Saving the dataset (158/192 shards): 82%\|████████▏ \| 260017/315947 [00:01<00:59, 944.80 examples/s] Sav
	3: ing the dataset (159/192 shards): 83%\|████████▎ \| 263307/315947 [00:01<00:55, 944.80 examples/s] Saving the dataset (160/192 shards): 83%\|████████▎ \| 263307/315947 [00:01<00:55, 944.80 examples/s] Saving the dataset (161/192 shards): 84%\|████████▍ \| 264952/315947 [00:01<00:53, 944.80 examples/s] Saving the dataset (162/192 shards): 84%\|████████▍ \| 266597/315947 [00:01<00:52, 944.80 examples/s] Saving the dataset (163/192 shards): 85%\|████████▍ \| 268242/315947 [00:01<00:50, 944.80 examples/s] Saving the dataset (164/192 shards): 86%\|████████▋ \| 273177/315947 [00:01<00:45, 944.80 examples/s] Saving the dataset (165/192 shards): 86%\|████████▋ \| 273177/315947 [00:01<00:45, 944.80 examples/s] Saving the dataset (166/192 shards): 87%\|████████▋ \| 274822/315947 [00:01<00:43, 944.80 examples/s] Saving the dataset (167/192 shards): 87%\|████████▋ \| 274
	3: 822/315947 [00:01<00:43, 944.80 examples/s] Saving the dataset (168/192 shards): 88%\|████████▊ \| 276467/315947 [00:01<00:41, 944.80 examples/s] Saving the dataset (169/192 shards): 88%\|████████▊ \| 278112/315947 [00:01<00:40, 944.80 examples/s] Saving the dataset (170/192 shards): 89%\|████████▉ \| 281402/315947 [00:01<00:36, 944.80 examples/s] Saving the dataset (171/192 shards): 89%\|████████▉ \| 281402/315947 [00:01<00:36, 944.80 examples/s] Saving the dataset (172/192 shards): 90%\|████████▉ \| 283047/315947 [00:01<00:34, 944.80 examples/s] Saving the dataset (173/192 shards): 91%\|█████████ \| 286337/315947 [00:01<00:31, 944.80 examples/s] Saving the dataset (174/192 shards): 91%\|█████████ \| 286337/315947 [00:01<00:31, 944.80 examples/s] Saving the dataset (175/192 shards): 91%\|█████████ \| 287982/315947 [00:01<00:29, 944.80 examples/s] Saving the dataset (176/192
	3: shards): 92%\|█████████▏\| 289627/315947 [00:01<00:27, 944.80 examples/s] Saving the dataset (177/192 shards): 92%\|█████████▏\| 291272/315947 [00:01<00:26, 944.80 examples/s] Saving the dataset (178/192 shards): 93%\|█████████▎\| 292917/315947 [00:01<00:24, 944.80 examples/s] Saving the dataset (179/192 shards): 93%\|█████████▎\| 294562/315947 [00:01<00:22, 944.80 examples/s] Saving the dataset (180/192 shards): 94%\|█████████▍\| 296207/315947 [00:01<00:20, 944.80 examples/s] Saving the dataset (181/192 shards): 94%\|█████████▍\| 297852/315947 [00:01<00:19, 944.80 examples/s] Saving the dataset (182/192 shards): 95%\|█████████▍\| 299497/315947 [00:01<00:17, 944.80 examples/s] Saving the dataset (183/192 shards): 95%\|█████████▌\| 301142/315947 [00:01<00:15, 944.80 examples/s] Saving the dataset (184/192 shards): 96%\|█████████▌\| 302787/315
	3: 947 [00:01<00:13, 944.80 examples/s] Saving the dataset (185/192 shards): 96%\|█████████▋\| 304432/315947 [00:01<00:12, 944.80 examples/s] Saving the dataset (186/192 shards): 97%\|█████████▋\| 306077/315947 [00:01<00:10, 944.80 examples/s] Saving the dataset (187/192 shards): 97%\|█████████▋\| 307722/315947 [00:01<00:08, 944.80 examples/s] Saving the dataset (188/192 shards): 98%\|█████████▊\| 309367/315947 [00:01<00:06, 944.80 examples/s] Saving the dataset (189/192 shards): 98%\|█████████▊\| 311012/315947 [00:01<00:05, 944.80 examples/s] Saving the dataset (190/192 shards): 99%\|█████████▉\| 312657/315947 [00:01<00:03, 944.80 examples/s] Saving the dataset (191/192 shards): 100%\|██████████\| 315947/315947 [00:01<00:00, 944.80 examples/s] Saving the dataset (192/192 shards): 100%\|██████████\| 315947/315947 [00:01<00:00, 944.80 examples/s] Saving the dataset
	3: (192/192 shards): 100%\|██████████\| 315947/315947 [00:01<00:00, 166814.88 examples/s]
	0: [2025-09-02 18:33:00,133] [INFO] [axolotl.utils.data.shared.load_preprocessed_dataset:472] [PID:801836] [RANK:0] Loading prepared dataset from disk at /lustre/fsn1/projects/rech/dgo/udv55np/dataset_math/Qwen3-235B-A22B/0/df4594deeb64977bb186d11562029017...[39m
	0: [2025-09-02 18:33:37,212] [INFO] [axolotl.utils.samplers.multipack.calc_min_len:436] [PID:801836] [RANK:0] gather_len_batches: [25939, 25940, 25939, 25938, 25938, 25939, 25939, 25939, 25939, 25938, 25939, 25939, 25939, 25939, 25939, 25938][39m
	0: [2025-09-02 18:33:37,417] [INFO] [axolotl.utils.trainer.calc_sample_packing_eff_est:495] [PID:801836] [RANK:0] sample_packing_eff_est across ranks: [0.9965550303459167, 0.9965550303459167, 0.9965550303459167, 0.9965550303459167, 0.9965166449546814, 0.9965550303459167, 0.996631920337677, 0.9965550303459167, 0.9965934753417969, 0.9965934753417969, 0.9965550303459167, 0.9965550303459167, 0.9965550303459167, 0.9965550303459167, 0.9965934753417969, 0.9965550303459167][39m
	0: [2025-09-02 18:33:37,422] [INFO] [axolotl.utils.data.sft._prepare_standard_dataset:127] [PID:801836] [RANK:0] Maximum number of steps set at 1621[39m
	0: [2025-09-02 18:33:37,749] [INFO] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_evaluation_loop:110] [PID:801836] [RANK:0] Patched Trainer.evaluation_loop with nanmean loss calculation[39m
	0: [2025-09-02 18:33:37,750] [INFO] [axolotl.monkeypatch.transformers.trainer_loss_calc.patch_maybe_log_save_evaluate:164] [PID:801836] [RANK:0] Patched Trainer._maybe_log_save_evaluate with nanmean loss calculation[39m
	0: [2025-09-02 18:33:50,442] [INFO] [axolotl.loaders.model._configure_embedding_dtypes:345] [PID:801836] [RANK:0] Converting modules to torch.bfloat16[39m
	0: [2025-09-02 18:34:01,367] [INFO] [axolotl.train.save_initial_configs:416] [PID:801836] [RANK:0] Pre-saving tokenizer to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0...[39m
	0: [2025-09-02 18:34:01,793] [INFO] [axolotl.train.save_initial_configs:419] [PID:801836] [RANK:0] Pre-saving model config to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0...[39m
	0: [2025-09-02 18:34:01,800] [INFO] [axolotl.train.execute_training:203] [PID:801836] [RANK:0] Starting trainer...[39m
	0: [2025-09-02 18:36:07,470] [INFO] [axolotl.utils.samplers.multipack.calc_min_len:436] [PID:801836] [RANK:0] gather_len_batches: [25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939, 25939][39m
	0: Parameter Offload - Persistent parameters statistics: param_count = 141, numel = 144896
	0: {'loss': 0.3291, 'grad_norm': 0.3215025832173914, 'learning_rate': 1.0779999999999999e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.01}
	0: 0%\| \| 0/1621 [00:00<?, ?it/s] 0%\| \| 1/1621 [03:15<87:56:27, 195.42s/it] 0%\| \| 2/1621 [03:18<36:53:28, 82.03s/it] 0%\| \| 3/1621 [03:18<20:12:28, 44.96s/it] 0%\| \| 4/1621 [03:19<12:22:43, 27.56s/it] 0%\| \| 5/1621 [03:20<8:03:17, 17.94s/it] 0%\| \| 6/1621 [03:21<5:26:43, 12.14s/it] 0%\| \| 7/1621 [03:22<3:47:13, 8.45s/it] 0%\| \| 8/1621 [03:23<2:42:07, 6.03s/it] 1%\| \| 9/1621 [03:24<1:58:26, 4.41s/it] 1%\| \| 10/1621 [03:24<1:29:01, 3.32s/it] 1%\| \| 10/1621 [03:25<1:29:01, 3.32s/it] 1%\| \| 11/1621 [03:25<1:09:22, 2.59s/it] 1%\| \| 12/1621 [03:26<55:27, 2.07s/it] 1%\| \| 13/1621 [03:27<47:03, 1.76s/it] 1%\| \| 14/1621 [03:28<39:48, 1.49s/it] 1%\| \| 15/1621 [03:29<34:37, 1.29s/it] 1%\| \| 16/1621 [03:30<31:04, 1.16s/it] 1%\| \| 17/1621 [03:31<28:54, 1.08s/it] 1%\|
	0: {'loss': 0.329, 'grad_norm': 0.2975306152955043, 'learning_rate': 1.498e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.01}
	0: {'loss': 0.332, 'grad_norm': 0.2988797728667972, 'learning_rate': 1.918e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.02}
	0: \| 18/1621 [03:32<28:10, 1.05s/it] 1%\| \| 19/1621 [03:33<26:28, 1.01it/s] 1%\| \| 20/1621 [03:34<25:36, 1.04it/s] 1%\| \| 20/1621 [03:34<25:36, 1.04it/s] 1%\|▏ \| 21/1621 [03:34<24:55, 1.07it/s] 1%\|▏ \| 22/1621 [03:35<24:19, 1.10it/s] 1%\|▏ \| 23/1621 [03:36<23:58, 1.11it/s] 1%\|▏ \| 24/1621 [03:37<23:28, 1.13it/s] 2%\|▏ \| 25/1621 [03:38<23:19, 1.14it/s] 2%\|▏ \| 26/1621 [03:39<23:58, 1.11it/s] 2%\|▏ \| 27/1621 [03:40<23:39, 1.12it/s] 2%\|▏ \| 28/1621 [03:41<23:17, 1.14it/s] 2%\|▏ \| 29/1621 [03:41<23:13, 1.14it/s] 2%\|▏ \| 30/1621 [03:42<23:56, 1.11it/s] 2%\|▏ \| 30/1621 [03:42<23:56, 1.11it/s] 2%\|▏ \| 31/1621 [03:43<23:34, 1.12it/s] 2%\|▏ \| 32/1621 [03:44<23:23, 1.13it/s] 2%\|▏ \| 33/1621 [03:45<23:01, 1.15it/s]
	0: {'loss': 0.3293, 'grad_norm': 0.31013818468501975, 'learning_rate': 2.338e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.02}
	0: 2%\|▏ \| 34/1621 [03:46<22:52, 1.16it/s] 2%\|▏ \| 35/1621 [03:47<22:41, 1.16it/s] 2%\|▏ \| 36/1621 [03:47<22:37, 1.17it/s] 2%\|▏ \| 37/1621 [03:48<22:30, 1.17it/s] 2%\|▏ \| 38/1621 [03:49<23:02, 1.14it/s] 2%\|▏ \| 39/1621 [03:50<22:53, 1.15it/s] 2%\|▏ \| 40/1621 [03:51<22:52, 1.15it/s] 2%\|▏ \| 40/1621 [03:51<22:52, 1.15it/s] 3%\|▎ \| 41/1621 [03:52<23:20, 1.13it/s] 3%\|▎ \| 42/1621 [03:53<23:09, 1.14it/s] 3%\|▎ \| 43/1621 [03:54<23:12, 1.13it/s] 3%\|▎ \| 44/1621 [03:54<23:03, 1.14it/s] 3%\|▎ \| 45/1621 [03:55<22:56, 1.15it/s] 3%\|▎ \| 46/1621 [03:56<22:47, 1.15it/s] 3%\|▎ \| 47/1621 [03:57<22:51, 1.15it/s] 3%\|▎ \| 48/1621 [03:58<22:36, 1.16it/s] 3%\|▎ \| 49/1621 [03:59<22:55, 1.14it/s] 3%\|▎ \| 50/1621 [04:00<23:14, 1.13it/s]
	0: {'loss': 0.3215, 'grad_norm': 0.30101641303027743, 'learning_rate': 2.758e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.03}
	0: {'loss': 0.3283, 'grad_norm': 0.28711923264082334, 'learning_rate': 3.1779999999999995e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.04}
	0: 3%\|▎ \| 50/1621 [04:00<23:14, 1.13it/s] 3%\|▎ \| 51/1621 [04:01<23:02, 1.14it/s] 3%\|▎ \| 52/1621 [04:01<22:45, 1.15it/s] 3%\|▎ \| 53/1621 [04:02<22:47, 1.15it/s] 3%\|▎ \| 54/1621 [04:03<22:50, 1.14it/s] 3%\|▎ \| 55/1621 [04:04<22:59, 1.14it/s] 3%\|▎ \| 56/1621 [04:05<22:56, 1.14it/s] 4%\|▎ \| 57/1621 [04:06<22:42, 1.15it/s] 4%\|▎ \| 58/1621 [04:07<22:38, 1.15it/s] 4%\|▎ \| 59/1621 [04:08<22:32, 1.15it/s] 4%\|▎ \| 60/1621 [04:08<22:22, 1.16it/s] 4%\|▎ \| 60/1621 [04:08<22:22, 1.16it/s] 4%\|▍ \| 61/1621 [04:09<23:12, 1.12it/s] 4%\|▍ \| 62/1621 [04:10<22:57, 1.13it/s] 4%\|▍ \| 63/1621 [04:11<22:45, 1.14it/s] 4%\|▍ \| 64/1621 [04:12<22:35, 1.15it/s] 4%\|▍ \| 65/1621 [04:13<23:04, 1.12it/s] 4%\|▍ \| 66/1621 [04:14<23:12, 1.12it/s] 4%\|▍ \| 67/
	0: {'loss': 0.3187, 'grad_norm': 0.3018654823455164, 'learning_rate': 3.598e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.04}
	0: {'loss': 0.3181, 'grad_norm': 0.2990949595796023, 'learning_rate': 4.0179999999999995e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.05}
	0: 1621 [04:15<22:58, 1.13it/s] 4%\|▍ \| 68/1621 [04:16<23:10, 1.12it/s] 4%\|▍ \| 69/1621 [04:17<24:41, 1.05it/s] 4%\|▍ \| 70/1621 [04:18<23:52, 1.08it/s] 4%\|▍ \| 70/1621 [04:18<23:52, 1.08it/s] 4%\|▍ \| 71/1621 [04:18<23:48, 1.09it/s] 4%\|▍ \| 72/1621 [04:19<23:25, 1.10it/s] 5%\|▍ \| 73/1621 [04:20<23:07, 1.12it/s] 5%\|▍ \| 74/1621 [04:21<22:42, 1.14it/s] 5%\|▍ \| 75/1621 [04:22<22:30, 1.15it/s] 5%\|▍ \| 76/1621 [04:23<22:22, 1.15it/s] 5%\|▍ \| 77/1621 [04:24<22:16, 1.16it/s] 5%\|▍ \| 78/1621 [04:24<22:13, 1.16it/s] 5%\|▍ \| 79/1621 [04:25<22:14, 1.16it/s] 5%\|▍ \| 80/1621 [04:26<23:04, 1.11it/s] 5%\|▍ \| 80/1621 [04:26<23:04, 1.11it/s] 5%\|▍ \| 81/1621 [04:27<23:33, 1.09it/s] 5%\|▌ \| 82/1621 [04:28<23:02, 1.11it/s] 5%\|▌
	0: {'loss': 0.318, 'grad_norm': 0.29019982626500473, 'learning_rate': 4.438e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.06}
	0: \| 83/1621 [04:29<22:51, 1.12it/s] 5%\|▌ \| 84/1621 [04:30<22:45, 1.13it/s] 5%\|▌ \| 85/1621 [04:31<22:46, 1.12it/s] 5%\|▌ \| 86/1621 [04:32<23:35, 1.08it/s] 5%\|▌ \| 87/1621 [04:33<23:29, 1.09it/s] 5%\|▌ \| 88/1621 [04:34<23:00, 1.11it/s] 5%\|▌ \| 89/1621 [04:34<22:40, 1.13it/s] 6%\|▌ \| 90/1621 [04:35<22:27, 1.14it/s] 6%\|▌ \| 90/1621 [04:35<22:27, 1.14it/s] 6%\|▌ \| 91/1621 [04:36<22:21, 1.14it/s] 6%\|▌ \| 92/1621 [04:37<22:09, 1.15it/s] 6%\|▌ \| 93/1621 [04:38<21:59, 1.16it/s] 6%\|▌ \| 94/1621 [04:39<21:56, 1.16it/s] 6%\|▌ \| 95/1621 [04:40<22:02, 1.15it/s] 6%\|▌ \| 96/1621 [04:40<21:59, 1.16it/s] 6%\|▌ \| 97/1621 [04:41<22:24, 1.13it/s] 6%\|▌ \| 98/1621 [04:42<22:17, 1.14it/s] 6%\|▌ \| 99/1621 [04:43<22:11, 1.14it/s] 6%\|▌ \| 100/1621 [04:44<21:54, 1.
	0: {'loss': 0.3195, 'grad_norm': 0.30163478964963014, 'learning_rate': 4.858000000000001e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.06}
	0: {'loss': 0.3174, 'grad_norm': 0.31809809662317745, 'learning_rate': 5.278e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.07}
	0: 16it/s] 6%\|▌ \| 100/1621 [04:44<21:54, 1.16it/s] 6%\|▌ \| 101/1621 [04:45<22:56, 1.10it/s] 6%\|▋ \| 102/1621 [04:46<22:32, 1.12it/s] 6%\|▋ \| 103/1621 [04:47<22:12, 1.14it/s] 6%\|▋ \| 104/1621 [04:47<22:01, 1.15it/s] 6%\|▋ \| 105/1621 [04:48<21:49, 1.16it/s] 7%\|▋ \| 106/1621 [04:49<21:45, 1.16it/s] 7%\|▋ \| 107/1621 [04:50<21:36, 1.17it/s] 7%\|▋ \| 108/1621 [04:51<21:34, 1.17it/s] 7%\|▋ \| 109/1621 [04:52<22:47, 1.11it/s] 7%\|▋ \| 110/1621 [04:53<22:25, 1.12it/s] 7%\|▋ \| 110/1621 [04:53<22:25, 1.12it/s] 7%\|▋ \| 111/1621 [04:54<22:11, 1.13it/s] 7%\|▋ \| 112/1621 [04:54<22:01, 1.14it/s] 7%\|▋ \| 113/1621 [04:55<21:55, 1.15it/s] 7%\|▋ \| 114/1621 [04:56<22:06, 1.14it/s] 7%\|▋ \| 115/1621 [04:57<21:55, 1.15it/s] 7%\|▋
	0: {'loss': 0.3304, 'grad_norm': 0.30854213193532004, 'learning_rate': 5.6979999999999995e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.07}
	0: {'loss': 0.32, 'grad_norm': 0.3074992484407681, 'learning_rate': 6.118e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.08}
	0: \| 116/1621 [04:58<21:48, 1.15it/s] 7%\|▋ \| 117/1621 [04:59<21:59, 1.14it/s] 7%\|▋ \| 118/1621 [05:00<21:48, 1.15it/s] 7%\|▋ \| 119/1621 [05:01<22:59, 1.09it/s] 7%\|▋ \| 120/1621 [05:02<22:38, 1.10it/s] 7%\|▋ \| 120/1621 [05:02<22:38, 1.10it/s] 7%\|▋ \| 121/1621 [05:02<22:11, 1.13it/s] 8%\|▊ \| 122/1621 [05:03<21:59, 1.14it/s] 8%\|▊ \| 123/1621 [05:04<21:46, 1.15it/s] 8%\|▊ \| 124/1621 [05:05<21:40, 1.15it/s] 8%\|▊ \| 125/1621 [05:06<21:33, 1.16it/s] 8%\|▊ \| 126/1621 [05:07<21:32, 1.16it/s] 8%\|▊ \| 127/1621 [05:08<21:53, 1.14it/s] 8%\|▊ \| 128/1621 [05:09<21:47, 1.14it/s] 8%\|▊ \| 129/1621 [05:09<21:36, 1.15it/s] 8%\|▊ \| 130/1621 [05:10<21:35, 1.15it/s] 8%\|▊ \| 130/1621 [05:10<21:35, 1.15it/s] 8%\|▊ \| 131/1621 [0
	0: {'loss': 0.3198, 'grad_norm': 0.3132505507666493, 'learning_rate': 6.538e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.09}
	0: 5:11<21:38, 1.15it/s] 8%\|▊ \| 132/1621 [05:12<21:26, 1.16it/s] 8%\|▊ \| 133/1621 [05:13<21:22, 1.16it/s] 8%\|▊ \| 134/1621 [05:14<21:17, 1.16it/s] 8%\|▊ \| 135/1621 [05:15<21:14, 1.17it/s] 8%\|▊ \| 136/1621 [05:15<21:12, 1.17it/s] 8%\|▊ \| 137/1621 [05:16<21:09, 1.17it/s] 9%\|▊ \| 138/1621 [05:17<21:05, 1.17it/s] 9%\|▊ \| 139/1621 [05:18<21:00, 1.18it/s] 9%\|▊ \| 140/1621 [05:19<21:00, 1.18it/s] 9%\|▊ \| 140/1621 [05:19<21:00, 1.18it/s] 9%\|▊ \| 141/1621 [05:20<21:16, 1.16it/s] 9%\|▉ \| 142/1621 [05:21<22:16, 1.11it/s] 9%\|▉ \| 143/1621 [05:22<21:49, 1.13it/s] 9%\|▉ \| 144/1621 [05:22<21:36, 1.14it/s] 9%\|▉ \| 145/1621 [05:23<21:27, 1.15it/s] 9%\|▉ \| 146/1621 [05:24<21:23, 1.15it/s] 9%\|▉ \| 147/1621 [05:25<21:28, 1.14it/s] 9%\|▉ \| 148/1621 [05:26<21:15, 1.15i
	0: {'loss': 0.3218, 'grad_norm': 0.3246947676119254, 'learning_rate': 6.958e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.09}
	0: {'loss': 0.3094, 'grad_norm': 0.28139341545138713, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.1}
	0: t/s] 9%\|▉ \| 149/1621 [05:27<21:22, 1.15it/s] 9%\|▉ \| 150/1621 [05:28<21:14, 1.15it/s] 9%\|▉ \| 150/1621 [05:28<21:14, 1.15it/s] 9%\|▉ \| 151/1621 [05:28<21:08, 1.16it/s] 9%\|▉ \| 152/1621 [05:29<22:18, 1.10it/s] 9%\|▉ \| 153/1621 [05:30<22:14, 1.10it/s] 10%\|▉ \| 154/1621 [05:31<22:51, 1.07it/s] 10%\|▉ \| 155/1621 [05:32<22:11, 1.10it/s] 10%\|▉ \| 156/1621 [05:33<21:56, 1.11it/s] 10%\|▉ \| 157/1621 [05:34<21:34, 1.13it/s] 10%\|▉ \| 158/1621 [05:35<21:22, 1.14it/s] 10%\|▉ \| 159/1621 [05:36<21:13, 1.15it/s] 10%\|▉ \| 160/1621 [05:37<21:05, 1.15it/s] 10%\|▉ \| 160/1621 [05:37<21:05, 1.15it/s] 10%\|▉ \| 161/1621 [05:37<21:02, 1.16it/s] 10%\|▉ \| 162/1621 [05:38<20:52, 1.16it/s] 10%\|█ \| 163/1621 [05:39<20:52, 1.16it/s] 10%\|█
	0: {'loss': 0.3101, 'grad_norm': 0.30704712278746976, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.1}
	0: \| 164/1621 [05:40<20:55, 1.16it/s] 10%\|█ \| 165/1621 [05:41<21:43, 1.12it/s] 10%\|█ \| 166/1621 [05:42<21:24, 1.13it/s] 10%\|█ \| 167/1621 [05:43<21:11, 1.14it/s] 10%\|█ \| 168/1621 [05:43<20:57, 1.16it/s] 10%\|█ \| 169/1621 [05:44<20:50, 1.16it/s] 10%\|█ \| 170/1621 [05:45<21:45, 1.11it/s] 10%\|█ \| 170/1621 [05:45<21:45, 1.11it/s] 11%\|█ \| 171/1621 [05:46<21:25, 1.13it/s] 11%\|█ \| 172/1621 [05:47<21:08, 1.14it/s] 11%\|█ \| 173/1621 [05:48<21:05, 1.14it/s] 11%\|█ \| 174/1621 [05:49<21:42, 1.11it/s] 11%\|█ \| 175/1621 [05:50<21:18, 1.13it/s] 11%\|█ \| 176/1621 [05:51<21:08, 1.14it/s] 11%\|█ \| 177/1621 [05:51<20:56, 1.15it/s] 11%\|█ \| 178/1621 [05:52<21:13, 1.13it/s] 11%\|█ \| 179/1621 [05:53<21:15, 1.13it/s] 11%\|█ \| 180/1621 [05:54<21:03, 1.14it/s]
	0: {'loss': 0.3146, 'grad_norm': 0.3287206705570871, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.11}
	0: {'loss': 0.3208, 'grad_norm': 0.3252333795917378, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.12}
	0: 11%\|█ \| 180/1621 [05:54<21:03, 1.14it/s] 11%\|█ \| 181/1621 [05:55<22:21, 1.07it/s] 11%\|█ \| 182/1621 [05:56<21:47, 1.10it/s] 11%\|█▏ \| 183/1621 [05:57<21:57, 1.09it/s] 11%\|█▏ \| 184/1621 [05:58<21:32, 1.11it/s] 11%\|█▏ \| 185/1621 [05:59<21:13, 1.13it/s] 11%\|█▏ \| 186/1621 [06:00<20:55, 1.14it/s] 12%\|█▏ \| 187/1621 [06:00<20:47, 1.15it/s] 12%\|█▏ \| 188/1621 [06:01<20:36, 1.16it/s] 12%\|█▏ \| 189/1621 [06:02<20:33, 1.16it/s] 12%\|█▏ \| 190/1621 [06:03<20:50, 1.14it/s] 12%\|█▏ \| 190/1621 [06:03<20:50, 1.14it/s] 12%\|█▏ \| 191/1621 [06:04<20:39, 1.15it/s] 12%\|█▏ \| 192/1621 [06:05<20:42, 1.15it/s] 12%\|█▏ \| 193/1621 [06:06<20:37, 1.15it/s] 12%\|█▏ \| 194/1621 [06:06<20:52, 1.14it/s] 12%\|█▏ \| 195/1621 [06:07<20:39, 1.15it/s] 12%\|█▏ \| 19
	0: {'loss': 0.3109, 'grad_norm': 0.3167312472541867, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.12}
	0: {'loss': 0.3198, 'grad_norm': 0.34127287905643694, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.13}
	0: 6/1621 [06:08<20:36, 1.15it/s] 12%\|█▏ \| 197/1621 [06:09<20:58, 1.13it/s] 12%\|█▏ \| 198/1621 [06:10<21:28, 1.10it/s] 12%\|█▏ \| 199/1621 [06:11<21:13, 1.12it/s] 12%\|█▏ \| 200/1621 [06:12<21:13, 1.12it/s] 12%\|█▏ \| 200/1621 [06:12<21:13, 1.12it/s] 12%\|█▏ \| 201/1621 [06:13<20:58, 1.13it/s] 12%\|█▏ \| 202/1621 [06:14<20:46, 1.14it/s] 13%\|█▎ \| 203/1621 [06:14<21:03, 1.12it/s] 13%\|█▎ \| 204/1621 [06:15<20:52, 1.13it/s] 13%\|█▎ \| 205/1621 [06:16<20:53, 1.13it/s] 13%\|█▎ \| 206/1621 [06:17<20:35, 1.15it/s] 13%\|█▎ \| 207/1621 [06:18<20:29, 1.15it/s] 13%\|█▎ \| 208/1621 [06:19<20:23, 1.15it/s] 13%\|█▎ \| 209/1621 [06:20<20:15, 1.16it/s] 13%\|█▎ \| 210/1621 [06:20<20:14, 1.16it/s] 13%\|█▎ \| 210/1621 [06:20<20:14, 1.16it/s] 13%\|█
	0: {'loss': 0.3123, 'grad_norm': 0.3156643302673956, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.14}
	0: ▎ \| 211/1621 [06:21<20:15, 1.16it/s] 13%\|█▎ \| 212/1621 [06:22<20:11, 1.16it/s] 13%\|█▎ \| 213/1621 [06:23<20:13, 1.16it/s] 13%\|█▎ \| 214/1621 [06:24<20:14, 1.16it/s] 13%\|█▎ \| 215/1621 [06:25<20:05, 1.17it/s] 13%\|█▎ \| 216/1621 [06:26<20:06, 1.16it/s] 13%\|█▎ \| 217/1621 [06:27<20:04, 1.17it/s] 13%\|█▎ \| 218/1621 [06:27<20:03, 1.17it/s] 14%\|█▎ \| 219/1621 [06:28<21:09, 1.10it/s] 14%\|█▎ \| 220/1621 [06:29<20:51, 1.12it/s] 14%\|█▎ \| 220/1621 [06:29<20:51, 1.12it/s] 14%\|█▎ \| 221/1621 [06:30<20:35, 1.13it/s] 14%\|█▎ \| 222/1621 [06:31<20:27, 1.14it/s] 14%\|█▍ \| 223/1621 [06:32<21:34, 1.08it/s] 14%\|█▍ \| 224/1621 [06:33<21:13, 1.10it/s] 14%\|█▍ \| 225/1621 [06:34<20:46, 1.12it/s] 14%\|█▍ \| 226/1621 [06:35<21:16, 1.09it/s] 14%\|█▍ \| 227/1621 [06:36<20:50,
	0: {'loss': 0.3132, 'grad_norm': 0.3383769344006066, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.14}
	0: {'loss': 0.3167, 'grad_norm': 0.31419076691138864, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.15}
	0: 1.12it/s] 14%\|█▍ \| 228/1621 [06:36<20:38, 1.12it/s] 14%\|█▍ \| 229/1621 [06:37<20:47, 1.12it/s] 14%\|█▍ \| 230/1621 [06:38<21:32, 1.08it/s] 14%\|█▍ \| 230/1621 [06:38<21:32, 1.08it/s] 14%\|█▍ \| 231/1621 [06:39<21:01, 1.10it/s] 14%\|█▍ \| 232/1621 [06:40<20:34, 1.13it/s] 14%\|█▍ \| 233/1621 [06:41<20:23, 1.13it/s] 14%\|█▍ \| 234/1621 [06:42<20:09, 1.15it/s] 14%\|█▍ \| 235/1621 [06:43<20:04, 1.15it/s] 15%\|█▍ \| 236/1621 [06:44<21:20, 1.08it/s] 15%\|█▍ \| 237/1621 [06:45<20:51, 1.11it/s] 15%\|█▍ \| 238/1621 [06:45<20:27, 1.13it/s] 15%\|█▍ \| 239/1621 [06:46<20:24, 1.13it/s] 15%\|█▍ \| 240/1621 [06:47<20:07, 1.14it/s] 15%\|█▍ \| 240/1621 [06:47<20:07, 1.14it/s] 15%\|█▍ \| 241/1621 [06:48<20:01, 1.15it/s] 15%\|█▍ \| 242/1621
	0: {'loss': 0.3046, 'grad_norm': 0.2902941154120115, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.15}
	0: [06:49<19:53, 1.16it/s] 15%\|█▍ \| 243/1621 [06:50<20:02, 1.15it/s] 15%\|█▌ \| 244/1621 [06:51<20:00, 1.15it/s] 15%\|█▌ \| 245/1621 [06:51<19:54, 1.15it/s] 15%\|█▌ \| 246/1621 [06:52<19:48, 1.16it/s] 15%\|█▌ \| 247/1621 [06:53<19:42, 1.16it/s] 15%\|█▌ \| 248/1621 [06:54<19:44, 1.16it/s] 15%\|█▌ \| 249/1621 [06:55<19:47, 1.16it/s] 15%\|█▌ \| 250/1621 [06:56<19:41, 1.16it/s] 15%\|█▌ \| 250/1621 [06:56<19:41, 1.16it/s] 15%\|█▌ \| 251/1621 [06:57<19:39, 1.16it/s] 16%\|█▌ \| 252/1621 [06:57<19:30, 1.17it/s] 16%\|█▌ \| 253/1621 [06:58<19:31, 1.17it/s] 16%\|█▌ \| 254/1621 [06:59<19:26, 1.17it/s] 16%\|█▌ \| 255/1621 [07:00<19:38, 1.16it/s] 16%\|█▌ \| 256/1621 [07:01<19:30, 1.17it/s] 16%\|█▌ \| 257/1621 [07:02<20:22, 1.12it/s] 16%\|█▌ \| 258/1621 [07:03<20:24, 1.11it/s] 16%\|█▌
	0: {'loss': 0.3108, 'grad_norm': 0.33968570459412806, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.16}
	0: {'loss': 0.3141, 'grad_norm': 0.3104904238715551, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.17}
	0: \| 259/1621 [07:04<20:18, 1.12it/s] 16%\|█▌ \| 260/1621 [07:05<20:03, 1.13it/s] 16%\|█▌ \| 260/1621 [07:05<20:03, 1.13it/s] 16%\|█▌ \| 261/1621 [07:05<19:53, 1.14it/s] 16%\|█▌ \| 262/1621 [07:06<19:44, 1.15it/s] 16%\|█▌ \| 263/1621 [07:07<19:33, 1.16it/s] 16%\|█▋ \| 264/1621 [07:08<19:29, 1.16it/s] 16%\|█▋ \| 265/1621 [07:09<20:03, 1.13it/s] 16%\|█▋ \| 266/1621 [07:10<19:52, 1.14it/s] 16%\|█▋ \| 267/1621 [07:11<20:18, 1.11it/s] 17%\|█▋ \| 268/1621 [07:12<20:00, 1.13it/s] 17%\|█▋ \| 269/1621 [07:13<20:30, 1.10it/s] 17%\|█▋ \| 270/1621 [07:13<20:13, 1.11it/s] 17%\|█▋ \| 270/1621 [07:13<20:13, 1.11it/s] 17%\|█▋ \| 271/1621 [07:14<19:53, 1.13it/s] 17%\|█▋ \| 272/1621 [07:15<20:32, 1.09it/s] 17%\|█▋ \| 273/1621 [07:16<20:58, 1.07it/
	0: {'loss': 0.3052, 'grad_norm': 0.29884303980196186, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.17}
	0: s] 17%\|█▋ \| 274/1621 [07:17<20:56, 1.07it/s] 17%\|█▋ \| 275/1621 [07:18<20:25, 1.10it/s] 17%\|█▋ \| 276/1621 [07:19<19:58, 1.12it/s] 17%\|█▋ \| 277/1621 [07:20<19:40, 1.14it/s] 17%\|█▋ \| 278/1621 [07:21<19:29, 1.15it/s] 17%\|█▋ \| 279/1621 [07:21<19:39, 1.14it/s] 17%\|█▋ \| 280/1621 [07:22<19:46, 1.13it/s] 17%\|█▋ \| 280/1621 [07:22<19:46, 1.13it/s] 17%\|█▋ \| 281/1621 [07:23<19:37, 1.14it/s] 17%\|█▋ \| 282/1621 [07:24<19:21, 1.15it/s] 17%\|█▋ \| 283/1621 [07:25<19:15, 1.16it/s] 18%\|█▊ \| 284/1621 [07:26<19:13, 1.16it/s] 18%\|█▊ \| 285/1621 [07:27<19:08, 1.16it/s] 18%\|█▊ \| 286/1621 [07:27<19:00, 1.17it/s] 18%\|█▊ \| 287/1621 [07:28<20:07, 1.10it/s] 18%\|█▊ \| 288/1621 [07:29<19:48, 1.12it/s] 18%\|█▊ \| 289/1621 [07:30<19:32, 1.14it/s] 18%\|█▊ \| 290/1621 [07:
	0: {'loss': 0.3096, 'grad_norm': 0.32007162423550123, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.18}
	0: {'loss': 0.3081, 'grad_norm': 0.36455586630760095, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.19}
	0: 31<19:16, 1.15it/s] 18%\|█▊ \| 290/1621 [07:31<19:16, 1.15it/s] 18%\|█▊ \| 291/1621 [07:32<19:11, 1.15it/s] 18%\|█▊ \| 292/1621 [07:33<19:07, 1.16it/s] 18%\|█▊ \| 293/1621 [07:34<19:04, 1.16it/s] 18%\|█▊ \| 294/1621 [07:34<19:10, 1.15it/s] 18%\|█▊ \| 295/1621 [07:35<19:11, 1.15it/s] 18%\|█▊ \| 296/1621 [07:36<19:03, 1.16it/s] 18%\|█▊ \| 297/1621 [07:37<19:06, 1.15it/s] 18%\|█▊ \| 298/1621 [07:38<19:35, 1.13it/s] 18%\|█▊ \| 299/1621 [07:39<19:20, 1.14it/s] 19%\|█▊ \| 300/1621 [07:40<19:09, 1.15it/s] 19%\|█▊ \| 300/1621 [07:40<19:09, 1.15it/s] 19%\|█▊ \| 301/1621 [07:41<19:04, 1.15it/s] 19%\|█▊ \| 302/1621 [07:41<18:57, 1.16it/s] 19%\|█▊ \| 303/1621 [07:42<18:56, 1.16it/s] 19%\|█▉ \| 304/1621 [07:43<19:05, 1.15it/s] 19%\|█▉
	0: {'loss': 0.309, 'grad_norm': 0.3243556638380136, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.19}
	0: {'loss': 0.3023, 'grad_norm': 0.3265912530967524, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.2}
	0: \| 305/1621 [07:44<19:03, 1.15it/s] 19%\|█▉ \| 306/1621 [07:45<18:58, 1.15it/s] 19%\|█▉ \| 307/1621 [07:46<18:53, 1.16it/s] 19%\|█▉ \| 308/1621 [07:47<19:06, 1.15it/s] 19%\|█▉ \| 309/1621 [07:48<18:52, 1.16it/s] 19%\|█▉ \| 310/1621 [07:48<18:46, 1.16it/s] 19%\|█▉ \| 310/1621 [07:48<18:46, 1.16it/s] 19%\|█▉ \| 311/1621 [07:49<19:02, 1.15it/s] 19%\|█▉ \| 312/1621 [07:50<18:50, 1.16it/s] 19%\|█▉ \| 313/1621 [07:51<18:59, 1.15it/s] 19%\|█▉ \| 314/1621 [07:52<19:00, 1.15it/s] 19%\|█▉ \| 315/1621 [07:53<19:26, 1.12it/s] 19%\|█▉ \| 316/1621 [07:54<19:19, 1.13it/s] 20%\|█▉ \| 317/1621 [07:55<19:05, 1.14it/s] 20%\|█▉ \| 318/1621 [07:55<19:22, 1.12it/s] 20%\|█▉ \| 319/1621 [07:56<19:11, 1.13it/s] 20%\|█▉ \| 320/1621 [07:57<19:16, 1.12it/s] 20%
	0: [2025-09-02 18:44:13,130] [INFO] [axolotl.core.trainers.base._save:613] [PID:801836] [RANK:0] Saving model checkpoint to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0/checkpoint-325[39m
	0: [2025-09-02 18:44:15,654] [INFO] [axolotl.core.trainers.base._save:662] [PID:801836] [RANK:0] Saving Trainer.data_collator.tokenizer by default as Trainer.processing_class is `None`[39m
	0: {'loss': 0.3072, 'grad_norm': 0.32927349219682783, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.2}
	0: \|█▉ \| 320/1621 [07:57<19:16, 1.12it/s] 20%\|█▉ \| 321/1621 [07:58<19:02, 1.14it/s] 20%\|█▉ \| 322/1621 [07:59<19:29, 1.11it/s] 20%\|█▉ \| 323/1621 [08:00<19:23, 1.12it/s] 20%\|█▉ \| 324/1621 [08:01<19:05, 1.13it/s] 20%\|██ \| 325/1621 [08:02<18:55, 1.14it/s] 20%\|██ \| 326/1621 [08:08<52:04, 2.41s/it] 20%\|██ \| 327/1621 [08:09<42:02, 1.95s/it] 20%\|██ \| 328/1621 [08:10<36:32, 1.70s/it] 20%\|██ \| 329/1621 [08:10<31:12, 1.45s/it] 20%\|██ \| 330/1621 [08:11<27:23, 1.27s/it] 20%\|██ \| 330/1621 [08:11<27:23, 1.27s/it] 20%\|██ \| 331/1621 [08:12<25:40, 1.19s/it] 20%\|██ \| 332/1621 [08:13<23:27, 1.09s/it] 21%\|██ \| 333/1621 [08:14<21:53, 1.02s/it] 21%\|██ \| 334/1621 [08:15<21:08, 1.01it/s] 21%\|██ \| 335/1621 [08:16<20:16, 1.06it/s] 21%\|██ \| 336/1621 [08:17<19:4
	0: {'loss': 0.3022, 'grad_norm': 0.3046050968707454, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.21}
	0: {'loss': 0.3001, 'grad_norm': 0.3260898742784537, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.22}
	0: 6, 1.08it/s] 21%\|██ \| 337/1621 [08:18<20:27, 1.05it/s] 21%\|██ \| 338/1621 [08:19<19:51, 1.08it/s] 21%\|██ \| 339/1621 [08:19<19:20, 1.10it/s] 21%\|██ \| 340/1621 [08:20<19:01, 1.12it/s] 21%\|██ \| 340/1621 [08:20<19:01, 1.12it/s] 21%\|██ \| 341/1621 [08:21<18:53, 1.13it/s] 21%\|██ \| 342/1621 [08:22<18:59, 1.12it/s] 21%\|██ \| 343/1621 [08:23<18:46, 1.13it/s] 21%\|██ \| 344/1621 [08:24<18:29, 1.15it/s] 21%\|██▏ \| 345/1621 [08:25<18:32, 1.15it/s] 21%\|██▏ \| 346/1621 [08:26<18:25, 1.15it/s] 21%\|██▏ \| 347/1621 [08:26<18:39, 1.14it/s] 21%\|██▏ \| 348/1621 [08:27<18:26, 1.15it/s] 22%\|██▏ \| 349/1621 [08:28<19:38, 1.08it/s] 22%\|██▏ \| 350/1621 [08:29<19:08, 1.11it/s] 22%\|██▏ \| 350/1621 [08:29<19:08, 1.11it/s] 22%\|██�
	0: {'loss': 0.3054, 'grad_norm': 0.3054229700445813, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.22}
	0: �� \| 351/1621 [08:30<18:50, 1.12it/s] 22%\|██▏ \| 352/1621 [08:31<18:33, 1.14it/s] 22%\|██▏ \| 353/1621 [08:32<18:27, 1.14it/s] 22%\|██▏ \| 354/1621 [08:33<18:21, 1.15it/s] 22%\|██▏ \| 355/1621 [08:33<18:12, 1.16it/s] 22%\|██▏ \| 356/1621 [08:34<18:14, 1.16it/s] 22%\|██▏ \| 357/1621 [08:35<18:10, 1.16it/s] 22%\|██▏ \| 358/1621 [08:36<18:01, 1.17it/s] 22%\|██▏ \| 359/1621 [08:37<17:57, 1.17it/s] 22%\|██▏ \| 360/1621 [08:38<18:23, 1.14it/s] 22%\|██▏ \| 360/1621 [08:38<18:23, 1.14it/s] 22%\|██▏ \| 361/1621 [08:39<18:30, 1.13it/s] 22%\|██▏ \| 362/1621 [08:40<18:19, 1.15it/s] 22%\|██▏ \| 363/1621 [08:40<18:08, 1.16it/s] 22%\|██▏ \| 364/1621 [08:41<18:06, 1.16it/s] 23%\|██▎ \| 365/1621 [08:42<18:02, 1.16it/s] 23%\|██▎ \| 366/1621 [08:43<17:59, 1.16it/s] 23%\|██▎
	0: {'loss': 0.3062, 'grad_norm': 0.3148357041588486, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.23}
	0: {'loss': 0.307, 'grad_norm': 0.31739638733721465, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.23}
	0: \| 367/1621 [08:44<18:03, 1.16it/s] 23%\|██▎ \| 368/1621 [08:45<18:00, 1.16it/s] 23%\|██▎ \| 369/1621 [08:46<18:14, 1.14it/s] 23%\|██▎ \| 370/1621 [08:46<18:03, 1.16it/s] 23%\|██▎ \| 370/1621 [08:46<18:03, 1.16it/s] 23%\|██▎ \| 371/1621 [08:47<17:58, 1.16it/s] 23%\|██▎ \| 372/1621 [08:48<18:06, 1.15it/s] 23%\|██▎ \| 373/1621 [08:49<17:58, 1.16it/s] 23%\|██▎ \| 374/1621 [08:50<18:02, 1.15it/s] 23%\|██▎ \| 375/1621 [08:51<17:52, 1.16it/s] 23%\|██▎ \| 376/1621 [08:52<18:30, 1.12it/s] 23%\|██▎ \| 377/1621 [08:53<18:33, 1.12it/s] 23%\|██▎ \| 378/1621 [08:53<18:19, 1.13it/s] 23%\|██▎ \| 379/1621 [08:54<18:23, 1.13it/s] 23%\|██▎ \| 380/1621 [08:55<18:05, 1.14it/s] 23%\|██▎ \| 380/1621 [08:55<18:05, 1.14it/s] 24%\|██▎ \|
	0: {'loss': 0.3102, 'grad_norm': 0.3375465905994638, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.24}
	0: 381/1621 [08:56<17:57, 1.15it/s] 24%\|██▎ \| 382/1621 [08:57<17:49, 1.16it/s] 24%\|██▎ \| 383/1621 [08:58<17:41, 1.17it/s] 24%\|██▎ \| 384/1621 [08:59<17:44, 1.16it/s] 24%\|██▍ \| 385/1621 [08:59<17:36, 1.17it/s] 24%\|██▍ \| 386/1621 [09:00<17:34, 1.17it/s] 24%\|██▍ \| 387/1621 [09:01<17:33, 1.17it/s] 24%\|██▍ \| 388/1621 [09:02<17:43, 1.16it/s] 24%\|██▍ \| 389/1621 [09:03<17:55, 1.15it/s] 24%\|██▍ \| 390/1621 [09:04<17:50, 1.15it/s] 24%\|██▍ \| 390/1621 [09:04<17:50, 1.15it/s] 24%\|██▍ \| 391/1621 [09:05<17:48, 1.15it/s] 24%\|██▍ \| 392/1621 [09:06<17:46, 1.15it/s] 24%\|██▍ \| 393/1621 [09:06<17:39, 1.16it/s] 24%\|██▍ \| 394/1621 [09:07<18:14, 1.12it/s] 24%\|██▍ \| 395/1621 [09:08<17:57, 1.14it/s] 24%\|██▍ \| 396/1621 [09:09<17:43, 1.15it/s] 24%\|██▍ \| 397
	0: {'loss': 0.3147, 'grad_norm': 0.34474370193771653, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.25}
	0: {'loss': 0.3014, 'grad_norm': 0.3087293542715505, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.25}
	0: /1621 [09:10<18:02, 1.13it/s] 25%\|██▍ \| 398/1621 [09:11<17:52, 1.14it/s] 25%\|██▍ \| 399/1621 [09:12<17:48, 1.14it/s] 25%\|██▍ \| 400/1621 [09:13<17:39, 1.15it/s] 25%\|██▍ \| 400/1621 [09:13<17:39, 1.15it/s] 25%\|██▍ \| 401/1621 [09:13<17:42, 1.15it/s] 25%\|██▍ \| 402/1621 [09:14<17:39, 1.15it/s] 25%\|██▍ \| 403/1621 [09:15<17:32, 1.16it/s] 25%\|██▍ \| 404/1621 [09:16<17:46, 1.14it/s] 25%\|██▍ \| 405/1621 [09:17<17:43, 1.14it/s] 25%\|██▌ \| 406/1621 [09:18<18:20, 1.10it/s] 25%\|██▌ \| 407/1621 [09:19<18:00, 1.12it/s] 25%\|██▌ \| 408/1621 [09:20<17:53, 1.13it/s] 25%\|██▌ \| 409/1621 [09:21<18:27, 1.09it/s] 25%\|██▌ \| 410/1621 [09:21<18:09, 1.11it/s] 25%\|██▌ \| 410/1621 [09:21<18:09, 1.11it/s] 25%\|██▌ \| 411/1621 [0
	0: {'loss': 0.3032, 'grad_norm': 0.3212422046835313, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.26}
	0: 9:22<17:56, 1.12it/s] 25%\|██▌ \| 412/1621 [09:23<17:48, 1.13it/s] 25%\|██▌ \| 413/1621 [09:24<17:41, 1.14it/s] 26%\|██▌ \| 414/1621 [09:25<17:29, 1.15it/s] 26%\|██▌ \| 415/1621 [09:26<17:32, 1.15it/s] 26%\|██▌ \| 416/1621 [09:27<18:28, 1.09it/s] 26%\|██▌ \| 417/1621 [09:28<18:01, 1.11it/s] 26%\|██▌ \| 418/1621 [09:29<17:48, 1.13it/s] 26%\|██▌ \| 419/1621 [09:29<17:33, 1.14it/s] 26%\|██▌ \| 420/1621 [09:30<17:30, 1.14it/s] 26%\|██▌ \| 420/1621 [09:30<17:30, 1.14it/s] 26%\|██▌ \| 421/1621 [09:31<17:23, 1.15it/s] 26%\|██▌ \| 422/1621 [09:32<17:15, 1.16it/s] 26%\|██▌ \| 423/1621 [09:33<17:11, 1.16it/s] 26%\|██▌ \| 424/1621 [09:34<17:06, 1.17it/s] 26%\|██▌ \| 425/1621 [09:35<17:34, 1.13it/s] 26%\|██▋ \| 426/1621 [09:36<17:35, 1.13it/s] 26%\|██▋ \| 427/1621 [09:3
	0: {'loss': 0.2992, 'grad_norm': 0.30649618094104975, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.27}
	0: {'loss': 0.3028, 'grad_norm': 0.3402735147626528, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.27}
	0: 7<18:16, 1.09it/s] 26%\|██▋ \| 428/1621 [09:37<17:54, 1.11it/s] 26%\|██▋ \| 429/1621 [09:38<17:34, 1.13it/s] 27%\|██▋ \| 430/1621 [09:39<17:23, 1.14it/s] 27%\|██▋ \| 430/1621 [09:39<17:23, 1.14it/s] 27%\|██▋ \| 431/1621 [09:40<17:11, 1.15it/s] 27%\|██▋ \| 432/1621 [09:41<17:06, 1.16it/s] 27%\|██▋ \| 433/1621 [09:42<17:14, 1.15it/s] 27%\|██▋ \| 434/1621 [09:43<17:36, 1.12it/s] 27%\|██▋ \| 435/1621 [09:43<17:31, 1.13it/s] 27%\|██▋ \| 436/1621 [09:44<17:19, 1.14it/s] 27%\|██▋ \| 437/1621 [09:45<17:16, 1.14it/s] 27%\|██▋ \| 438/1621 [09:46<17:21, 1.14it/s] 27%\|██▋ \| 439/1621 [09:47<17:29, 1.13it/s] 27%\|██▋ \| 440/1621 [09:48<17:14, 1.14it/s] 27%\|██▋ \| 440/1621 [09:48<17:14, 1.14it/s] 27%\|██▋ \| 441/1621 [09:49<17:08,
	0: {'loss': 0.3009, 'grad_norm': 0.30752642276161624, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.28}
	0: 1.15it/s] 27%\|██▋ \| 442/1621 [09:50<17:03, 1.15it/s] 27%\|██▋ \| 443/1621 [09:50<16:57, 1.16it/s] 27%\|██▋ \| 444/1621 [09:51<16:55, 1.16it/s] 27%\|██▋ \| 445/1621 [09:52<16:48, 1.17it/s] 28%\|██▊ \| 446/1621 [09:53<16:54, 1.16it/s] 28%\|██▊ \| 447/1621 [09:54<16:47, 1.16it/s] 28%\|██▊ \| 448/1621 [09:55<16:52, 1.16it/s] 28%\|██▊ \| 449/1621 [09:56<16:46, 1.16it/s] 28%\|██▊ \| 450/1621 [09:56<17:01, 1.15it/s] 28%\|██▊ \| 450/1621 [09:56<17:01, 1.15it/s] 28%\|██▊ \| 451/1621 [09:57<16:53, 1.15it/s] 28%\|██▊ \| 452/1621 [09:58<16:46, 1.16it/s] 28%\|██▊ \| 453/1621 [09:59<16:50, 1.16it/s] 28%\|██▊ \| 454/1621 [10:00<17:19, 1.12it/s] 28%\|██▊ \| 455/1621 [10:01<17:11, 1.13it/s] 28%\|██▊ \| 456/1621 [10:02<16:57, 1.15it/s] 28%\|██▊ \| 457/1621 [10:03<16:49, 1
	0: {'loss': 0.2975, 'grad_norm': 0.2988914632135753, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.28}
	0: {'loss': 0.3047, 'grad_norm': 0.3029109045844161, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.29}
	0: .15it/s] 28%\|██▊ \| 458/1621 [10:03<16:52, 1.15it/s] 28%\|██▊ \| 459/1621 [10:04<16:46, 1.15it/s] 28%\|██▊ \| 460/1621 [10:05<16:49, 1.15it/s] 28%\|██▊ \| 460/1621 [10:05<16:49, 1.15it/s] 28%\|██▊ \| 461/1621 [10:06<17:07, 1.13it/s] 29%\|██▊ \| 462/1621 [10:07<17:13, 1.12it/s] 29%\|██▊ \| 463/1621 [10:08<17:41, 1.09it/s] 29%\|██▊ \| 464/1621 [10:09<17:19, 1.11it/s] 29%\|██▊ \| 465/1621 [10:10<17:01, 1.13it/s] 29%\|██▊ \| 466/1621 [10:11<18:08, 1.06it/s] 29%\|██▉ \| 467/1621 [10:12<17:39, 1.09it/s] 29%\|██▉ \| 468/1621 [10:13<17:26, 1.10it/s] 29%\|██▉ \| 469/1621 [10:13<17:14, 1.11it/s] 29%\|██▉ \| 470/1621 [10:14<16:54, 1.13it/s] 29%\|██▉ \| 470/1621 [10:14<16:54, 1.13it/s] 29%\|██▉ \| 471/1621 [10:15<16:47, 1.14it/s]
	0: {'loss': 0.304, 'grad_norm': 0.30933769716252907, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.3}
	0: 29%\|██▉ \| 472/1621 [10:16<16:55, 1.13it/s] 29%\|██▉ \| 473/1621 [10:17<17:04, 1.12it/s] 29%\|██▉ \| 474/1621 [10:18<16:49, 1.14it/s] 29%\|██▉ \| 475/1621 [10:19<17:08, 1.11it/s] 29%\|██▉ \| 476/1621 [10:20<17:00, 1.12it/s] 29%\|██▉ \| 477/1621 [10:20<16:47, 1.14it/s] 29%\|██▉ \| 478/1621 [10:21<16:36, 1.15it/s] 30%\|██▉ \| 479/1621 [10:22<16:41, 1.14it/s] 30%\|██▉ \| 480/1621 [10:23<16:35, 1.15it/s] 30%\|██▉ \| 480/1621 [10:23<16:35, 1.15it/s] 30%\|██▉ \| 481/1621 [10:24<17:18, 1.10it/s] 30%\|██▉ \| 482/1621 [10:25<16:58, 1.12it/s] 30%\|██▉ \| 483/1621 [10:26<16:58, 1.12it/s] 30%\|██▉ \| 484/1621 [10:27<16:43, 1.13it/s] 30%\|██▉ \| 485/1621 [10:28<17:17, 1.09it/s] 30%\|██▉ \| 486/1621 [10:28<16:54, 1.12it/s] 30%\|███ \| 487/1621 [10:29<17:25, 1.08it/s] 3
	0: {'loss': 0.304, 'grad_norm': 0.3140435122849643, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.3}
	0: {'loss': 0.3019, 'grad_norm': 0.299751756066974, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.31}
	0: 0%\|███ \| 488/1621 [10:30<17:02, 1.11it/s] 30%\|███ \| 489/1621 [10:31<16:43, 1.13it/s] 30%\|███ \| 490/1621 [10:32<16:30, 1.14it/s] 30%\|███ \| 490/1621 [10:32<16:30, 1.14it/s] 30%\|███ \| 491/1621 [10:33<16:56, 1.11it/s] 30%\|███ \| 492/1621 [10:34<16:38, 1.13it/s] 30%\|███ \| 493/1621 [10:35<16:24, 1.15it/s] 30%\|███ \| 494/1621 [10:36<16:32, 1.14it/s] 31%\|███ \| 495/1621 [10:36<16:26, 1.14it/s] 31%\|███ \| 496/1621 [10:37<16:31, 1.13it/s] 31%\|███ \| 497/1621 [10:38<17:00, 1.10it/s] 31%\|███ \| 498/1621 [10:39<16:40, 1.12it/s] 31%\|███ \| 499/1621 [10:40<16:25, 1.14it/s] 31%\|███ \| 500/1621 [10:41<16:18, 1.15it/s] 31%\|███ \| 500/1621 [10:41<16:18, 1.15it/s] 31%\|███ \| 501/1621 [10:42<16:15, 1.15it/s] 31%\|█�
	0: {'loss': 0.3022, 'grad_norm': 0.31890034279759916, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.31}
	0: �█ \| 502/1621 [10:43<16:09, 1.15it/s] 31%\|███ \| 503/1621 [10:43<16:21, 1.14it/s] 31%\|███ \| 504/1621 [10:44<16:09, 1.15it/s] 31%\|███ \| 505/1621 [10:45<16:05, 1.16it/s] 31%\|███ \| 506/1621 [10:46<16:57, 1.10it/s] 31%\|███▏ \| 507/1621 [10:47<16:51, 1.10it/s] 31%\|███▏ \| 508/1621 [10:48<16:33, 1.12it/s] 31%\|███▏ \| 509/1621 [10:49<16:17, 1.14it/s] 31%\|███▏ \| 510/1621 [10:50<16:03, 1.15it/s] 31%\|███▏ \| 510/1621 [10:50<16:03, 1.15it/s] 32%\|███▏ \| 511/1621 [10:51<16:37, 1.11it/s] 32%\|███▏ \| 512/1621 [10:52<16:27, 1.12it/s] 32%\|███▏ \| 513/1621 [10:52<16:36, 1.11it/s] 32%\|███▏ \| 514/1621 [10:53<16:20, 1.13it/s] 32%\|███▏ \| 515/1621 [10:54<16:25, 1.12it/s] 32%\|███▏ \| 516/1621 [10:55<16:12, 1.14it/s] 32%\|███▏ \| 517/1621 [10:56<16:00,
	0: {'loss': 0.3021, 'grad_norm': 0.31952587601461685, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.32}
	0: {'loss': 0.2955, 'grad_norm': 0.3094649616845461, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.33}
	0: 1.15it/s] 32%\|███▏ \| 518/1621 [10:57<16:01, 1.15it/s] 32%\|███▏ \| 519/1621 [10:58<17:06, 1.07it/s] 32%\|███▏ \| 520/1621 [10:59<16:49, 1.09it/s] 32%\|███▏ \| 520/1621 [10:59<16:49, 1.09it/s] 32%\|███▏ \| 521/1621 [11:00<16:36, 1.10it/s] 32%\|███▏ \| 522/1621 [11:01<16:48, 1.09it/s] 32%\|███▏ \| 523/1621 [11:01<16:30, 1.11it/s] 32%\|███▏ \| 524/1621 [11:02<17:10, 1.06it/s] 32%\|███▏ \| 525/1621 [11:03<16:50, 1.08it/s] 32%\|███▏ \| 526/1621 [11:04<16:42, 1.09it/s] 33%\|███▎ \| 527/1621 [11:05<16:49, 1.08it/s] 33%\|███▎ \| 528/1621 [11:06<16:41, 1.09it/s] 33%\|███▎ \| 529/1621 [11:07<16:28, 1.10it/s] 33%\|███▎ \| 530/1621 [11:08<16:13, 1.12it/s] 33%\|███▎ \| 530/1621 [11:08<16:13, 1.12it/s] 33%\|███▎ \|
	0: {'loss': 0.2986, 'grad_norm': 0.3112632339780639, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.33}
	0: 531/1621 [11:09<16:42, 1.09it/s] 33%\|███▎ \| 532/1621 [11:10<16:54, 1.07it/s] 33%\|███▎ \| 533/1621 [11:11<16:29, 1.10it/s] 33%\|███▎ \| 534/1621 [11:11<16:16, 1.11it/s] 33%\|███▎ \| 535/1621 [11:12<16:01, 1.13it/s] 33%\|███▎ \| 536/1621 [11:13<16:04, 1.13it/s] 33%\|███▎ \| 537/1621 [11:14<15:51, 1.14it/s] 33%\|███▎ \| 538/1621 [11:15<15:40, 1.15it/s] 33%\|███▎ \| 539/1621 [11:16<15:32, 1.16it/s] 33%\|███▎ \| 540/1621 [11:17<15:36, 1.15it/s] 33%\|███▎ \| 540/1621 [11:17<15:36, 1.15it/s] 33%\|███▎ \| 541/1621 [11:17<15:29, 1.16it/s] 33%\|███▎ \| 542/1621 [11:18<15:51, 1.13it/s] 33%\|███▎ \| 543/1621 [11:19<15:49, 1.14it/s] 34%\|███▎ \| 544/1621 [11:20<15:40, 1.15it/s] 34%\|███▎ \| 545/1621 [11:21<16:22, 1.09it/s] 34%\|███▎ \| 546/1621 [11:22<16:06, 1.1
	0: {'loss': 0.306, 'grad_norm': 0.32623249352975026, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.34}
	0: {'loss': 0.303, 'grad_norm': 0.3212950878739604, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.35}
	0: 1it/s] 34%\|███▎ \| 547/1621 [11:23<15:54, 1.13it/s] 34%\|███▍ \| 548/1621 [11:24<15:43, 1.14it/s] 34%\|███▍ \| 549/1621 [11:25<15:45, 1.13it/s] 34%\|███▍ \| 550/1621 [11:25<15:33, 1.15it/s] 34%\|███▍ \| 550/1621 [11:25<15:33, 1.15it/s] 34%\|███▍ \| 551/1621 [11:26<15:27, 1.15it/s] 34%\|███▍ \| 552/1621 [11:27<15:29, 1.15it/s] 34%\|███▍ \| 553/1621 [11:28<15:23, 1.16it/s] 34%\|███▍ \| 554/1621 [11:29<15:26, 1.15it/s] 34%\|███▍ \| 555/1621 [11:30<15:21, 1.16it/s] 34%\|███▍ \| 556/1621 [11:31<16:22, 1.08it/s] 34%\|███▍ \| 557/1621 [11:32<16:02, 1.11it/s] 34%\|███▍ \| 558/1621 [11:33<15:53, 1.12it/s] 34%\|███▍ \| 559/1621 [11:34<16:40, 1.06it/s] 35%\|███▍ \| 560/1621 [11:34<16:09, 1.09it/s] 35%\|███▍ \| 560
	0: {'loss': 0.2953, 'grad_norm': 0.2767562705424428, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.35}
	0: /1621 [11:35<16:09, 1.09it/s] 35%\|███▍ \| 561/1621 [11:35<15:52, 1.11it/s] 35%\|███▍ \| 562/1621 [11:36<15:35, 1.13it/s] 35%\|███▍ \| 563/1621 [11:37<15:25, 1.14it/s] 35%\|███▍ \| 564/1621 [11:38<15:17, 1.15it/s] 35%\|███▍ \| 565/1621 [11:39<15:18, 1.15it/s] 35%\|███▍ \| 566/1621 [11:40<15:10, 1.16it/s] 35%\|███▍ \| 567/1621 [11:40<15:06, 1.16it/s] 35%\|███▌ \| 568/1621 [11:41<15:03, 1.16it/s] 35%\|███▌ \| 569/1621 [11:42<15:55, 1.10it/s] 35%\|███▌ \| 570/1621 [11:43<15:38, 1.12it/s] 35%\|███▌ \| 570/1621 [11:43<15:38, 1.12it/s] 35%\|███▌ \| 571/1621 [11:44<15:26, 1.13it/s] 35%\|███▌ \| 572/1621 [11:45<15:14, 1.15it/s] 35%\|███▌ \| 573/1621 [11:46<15:08, 1.15it/s] 35%\|███▌ \| 574/1621 [11:47<15:01, 1.16it/s] 35%\|███▌ \| 575/1621 [11:47<14:59, 1.16it/
	0: {'loss': 0.2946, 'grad_norm': 0.285031036391949, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.36}
	0: s] 36%\|███▌ \| 576/1621 [11:48<15:28, 1.13it/s] 36%\|███▌ \| 577/1621 [11:49<15:16, 1.14it/s] 36%\|███▌ \| 578/1621 [11:50<15:21, 1.13it/s] 36%\|███▌ \| 579/1621 [11:51<15:13, 1.14it/s] 36%\|███▌ \| 580/1621 [11:52<15:05, 1.15it/s] 36%\|███▌ \| 580/1621 [11:52<15:05, 1.15it/s] 36%\|███▌ \| 581/1621 [11:53<15:31, 1.12it/s] 36%\|███▌ \| 582/1621 [11:54<15:17, 1.13it/s] 36%\|███▌ \| 583/1621 [11:55<16:08, 1.07it/s] 36%\|███▌ \| 584/1621 [11:56<15:40, 1.10it/s] 36%\|███▌ \| 585/1621 [11:56<15:23, 1.12it/s] 36%\|███▌ \| 586/1621 [11:57<15:09, 1.14it/s] 36%\|███▌ \| 587/1621 [11:58<15:01, 1.15it/s] 36%\|███▋ \| 588/1621 [11:59<14:52, 1.16it/s] 36%\|███▋ \| 589/1621 [12:00<14:53, 1.15it/s] 36%\|███▋ \| 590/1621 [12:01<15:04, 1.14it/s]
	0: {'loss': 0.2978, 'grad_norm': 0.3116693864450672, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.36}
	0: {'loss': 0.2942, 'grad_norm': 0.29553887428879677, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.37}
	0: 36%\|███▋ \| 590/1621 [12:01<15:04, 1.14it/s] 36%\|███▋ \| 591/1621 [12:02<14:56, 1.15it/s] 37%\|███▋ \| 592/1621 [12:03<14:55, 1.15it/s] 37%\|███▋ \| 593/1621 [12:03<14:58, 1.14it/s] 37%\|███▋ \| 594/1621 [12:04<14:50, 1.15it/s] 37%\|███▋ \| 595/1621 [12:05<15:31, 1.10it/s] 37%\|███▋ \| 596/1621 [12:06<15:20, 1.11it/s] 37%\|███▋ \| 597/1621 [12:07<15:14, 1.12it/s] 37%\|███▋ \| 598/1621 [12:08<15:00, 1.14it/s] 37%\|███▋ \| 599/1621 [12:09<15:04, 1.13it/s] 37%\|███▋ \| 600/1621 [12:10<15:11, 1.12it/s] 37%\|███▋ \| 600/1621 [12:10<15:11, 1.12it/s] 37%\|███▋ \| 601/1621 [12:11<14:59, 1.13it/s] 37%\|███▋ \| 602/1621 [12:11<14:57, 1.14it/s] 37%\|███▋ \| 603/1621 [12:12<14:57, 1.13it/s] 37%\|███▋ \| 604/1621 [12:13<15:10, 1.12it/s]
	0: {'loss': 0.2993, 'grad_norm': 0.32060551637366747, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.38}
	0: 37%\|███▋ \| 605/1621 [12:14<14:53, 1.14it/s] 37%\|███▋ \| 606/1621 [12:15<14:59, 1.13it/s] 37%\|███▋ \| 607/1621 [12:16<14:53, 1.13it/s] 38%\|███▊ \| 608/1621 [12:17<15:27, 1.09it/s] 38%\|███▊ \| 609/1621 [12:18<15:33, 1.08it/s] 38%\|███▊ \| 610/1621 [12:19<15:09, 1.11it/s] 38%\|███▊ \| 610/1621 [12:19<15:09, 1.11it/s] 38%\|███▊ \| 611/1621 [12:20<15:39, 1.08it/s] 38%\|███▊ \| 612/1621 [12:20<15:20, 1.10it/s] 38%\|███▊ \| 613/1621 [12:21<15:12, 1.11it/s] 38%\|███▊ \| 614/1621 [12:22<14:57, 1.12it/s] 38%\|███▊ \| 615/1621 [12:23<15:13, 1.10it/s] 38%\|███▊ \| 616/1621 [12:24<16:41, 1.00it/s] 38%\|███▊ \| 617/1621 [12:25<15:55, 1.05it/s] 38%\|███▊ \| 618/1621 [12:26<15:27, 1.08it/s] 38%\|███▊ \| 619/1621 [12:27<15:43, 1.06it/s] 38%\|███▊ \| 620/
	0: {'loss': 0.2881, 'grad_norm': 0.3005663461419241, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.38}
	0: {'loss': 0.2984, 'grad_norm': 0.30549174693463893, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.39}
	0: 1621 [12:28<15:19, 1.09it/s] 38%\|███▊ \| 620/1621 [12:28<15:19, 1.09it/s] 38%\|███▊ \| 621/1621 [12:29<15:01, 1.11it/s] 38%\|███▊ \| 622/1621 [12:30<14:45, 1.13it/s] 38%\|███▊ \| 623/1621 [12:30<14:32, 1.14it/s] 38%\|███▊ \| 624/1621 [12:31<14:27, 1.15it/s] 39%\|███▊ \| 625/1621 [12:32<14:20, 1.16it/s] 39%\|███▊ \| 626/1621 [12:33<15:02, 1.10it/s] 39%\|███▊ \| 627/1621 [12:34<14:42, 1.13it/s] 39%\|███▊ \| 628/1621 [12:35<15:30, 1.07it/s] 39%\|███▉ \| 629/1621 [12:36<15:40, 1.05it/s] 39%\|███▉ \| 630/1621 [12:37<15:18, 1.08it/s] 39%\|███▉ \| 630/1621 [12:37<15:18, 1.08it/s] 39%\|███▉ \| 631/1621 [12:38<15:01, 1.10it/s] 39%\|███▉ \| 632/1621 [12:39<14:59, 1.10it/s] 39%\|███▉ \| 633/1621 [12:40<14:41, 1.12it/s] 39%\|
	0: {'loss': 0.2927, 'grad_norm': 0.3779583398963665, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.39}
	0: ███▉ \| 634/1621 [12:41<15:25, 1.07it/s] 39%\|███▉ \| 635/1621 [12:42<15:20, 1.07it/s] 39%\|███▉ \| 636/1621 [12:42<14:59, 1.10it/s] 39%\|███▉ \| 637/1621 [12:43<14:49, 1.11it/s] 39%\|███▉ \| 638/1621 [12:44<14:36, 1.12it/s] 39%\|███▉ \| 639/1621 [12:45<14:31, 1.13it/s] 39%\|███▉ \| 640/1621 [12:46<14:52, 1.10it/s] 39%\|███▉ \| 640/1621 [12:46<14:52, 1.10it/s] 40%\|███▉ \| 641/1621 [12:47<15:10, 1.08it/s] 40%\|███▉ \| 642/1621 [12:48<15:05, 1.08it/s] 40%\|███▉ \| 643/1621 [12:49<14:44, 1.11it/s] 40%\|███▉ \| 644/1621 [12:50<14:25, 1.13it/s] 40%\|███▉ \| 645/1621 [12:50<14:16, 1.14it/s] 40%\|███▉ \| 646/1621 [12:51<14:13, 1.14it/s] 40%\|███▉ \| 647/1621 [12:52<14:06, 1.15it/s] 40%\|███▉ \| 648/1621 [12:53<14:03, 1.15it/s] 40%\|████ \| 649/1621
	0: {'loss': 0.2996, 'grad_norm': 0.29989530504483425, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.4}
	0: [2025-09-02 18:49:05,656] [INFO] [axolotl.core.trainers.base._save:613] [PID:801836] [RANK:0] Saving model checkpoint to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0/checkpoint-650[39m
	0: [2025-09-02 18:49:08,181] [INFO] [axolotl.core.trainers.base._save:662] [PID:801836] [RANK:0] Saving Trainer.data_collator.tokenizer by default as Trainer.processing_class is `None`[39m
	0: {'loss': 0.2977, 'grad_norm': 0.30302027522488867, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.41}
	0: [12:54<13:56, 1.16it/s] 40%\|████ \| 650/1621 [12:55<13:52, 1.17it/s] 40%\|████ \| 650/1621 [12:55<13:52, 1.17it/s] 40%\|████ \| 651/1621 [13:00<35:49, 2.22s/it] 40%\|████ \| 652/1621 [13:01<29:12, 1.81s/it] 40%\|████ \| 653/1621 [13:02<24:30, 1.52s/it] 40%\|████ \| 654/1621 [13:03<21:21, 1.33s/it] 40%\|████ \| 655/1621 [13:04<19:03, 1.18s/it] 40%\|████ \| 656/1621 [13:04<17:25, 1.08s/it] 41%\|████ \| 657/1621 [13:05<16:16, 1.01s/it] 41%\|████ \| 658/1621 [13:06<15:29, 1.04it/s] 41%\|████ \| 659/1621 [13:07<15:09, 1.06it/s] 41%\|████ \| 660/1621 [13:08<14:41, 1.09it/s] 41%\|████ \| 660/1621 [13:08<14:41, 1.09it/s] 41%\|████ \| 661/1621 [13:09<14:27, 1.11it/s] 41%\|████ \| 662/1621 [13:10<14:10, 1.13it/s] 41%\|█�
	0: {'loss': 0.2976, 'grad_norm': 0.2892331257316936, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.41}
	0: ��██ \| 663/1621 [13:10<14:01, 1.14it/s] 41%\|████ \| 664/1621 [13:11<14:20, 1.11it/s] 41%\|████ \| 665/1621 [13:12<14:03, 1.13it/s] 41%\|████ \| 666/1621 [13:13<14:12, 1.12it/s] 41%\|████ \| 667/1621 [13:14<14:02, 1.13it/s] 41%\|████ \| 668/1621 [13:15<13:59, 1.14it/s] 41%\|████▏ \| 669/1621 [13:16<13:52, 1.14it/s] 41%\|████▏ \| 670/1621 [13:17<13:46, 1.15it/s] 41%\|████▏ \| 670/1621 [13:17<13:46, 1.15it/s] 41%\|████▏ \| 671/1621 [13:18<14:02, 1.13it/s] 41%\|████▏ \| 672/1621 [13:18<13:54, 1.14it/s] 42%\|████▏ \| 673/1621 [13:19<13:46, 1.15it/s] 42%\|████▏ \| 674/1621 [13:20<13:38, 1.16it/s] 42%\|████▏ \| 675/1621 [13:21<13:34, 1.16it/s] 42%\|████▏ \| 676/1621 [13:22<13:38, 1.15it/s] 42%\|████▏ \| 677/1621 [13:23<13:34, 1.16it/s] 42%\|████
	0: {'loss': 0.2914, 'grad_norm': 0.30196767512453576, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.42}
	0: {'loss': 0.2932, 'grad_norm': 0.29700083045306425, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.43}
	0: ▏ \| 678/1621 [13:24<14:00, 1.12it/s] 42%\|████▏ \| 679/1621 [13:25<14:12, 1.10it/s] 42%\|████▏ \| 680/1621 [13:25<13:56, 1.13it/s] 42%\|████▏ \| 680/1621 [13:25<13:56, 1.13it/s] 42%\|████▏ \| 681/1621 [13:26<13:44, 1.14it/s] 42%\|████▏ \| 682/1621 [13:27<13:34, 1.15it/s] 42%\|████▏ \| 683/1621 [13:28<13:33, 1.15it/s] 42%\|████▏ \| 684/1621 [13:29<13:38, 1.14it/s] 42%\|████▏ \| 685/1621 [13:30<13:34, 1.15it/s] 42%\|████▏ \| 686/1621 [13:31<13:52, 1.12it/s] 42%\|████▏ \| 687/1621 [13:32<13:38, 1.14it/s] 42%\|████▏ \| 688/1621 [13:33<14:33, 1.07it/s] 43%\|████▎ \| 689/1621 [13:33<14:07, 1.10it/s] 43%\|████▎ \| 690/1621 [13:34<14:09, 1.10it/s] 43%\|████▎ \| 690/1621 [13:34<14:09, 1.10it/s] 43%\|████▎
	0: {'loss': 0.3008, 'grad_norm': 0.2990684455792348, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.43}
	0: \| 691/1621 [13:35<13:51, 1.12it/s] 43%\|████▎ \| 692/1621 [13:36<13:45, 1.13it/s] 43%\|████▎ \| 693/1621 [13:37<13:32, 1.14it/s] 43%\|████▎ \| 694/1621 [13:38<13:25, 1.15it/s] 43%\|████▎ \| 695/1621 [13:39<13:19, 1.16it/s] 43%\|████▎ \| 696/1621 [13:39<13:12, 1.17it/s] 43%\|████▎ \| 697/1621 [13:40<13:16, 1.16it/s] 43%\|████▎ \| 698/1621 [13:41<13:16, 1.16it/s] 43%\|████▎ \| 699/1621 [13:42<13:13, 1.16it/s] 43%\|████▎ \| 700/1621 [13:43<13:10, 1.17it/s] 43%\|████▎ \| 700/1621 [13:43<13:10, 1.17it/s] 43%\|████▎ \| 701/1621 [13:44<13:20, 1.15it/s] 43%\|████▎ \| 702/1621 [13:45<13:15, 1.16it/s] 43%\|████▎ \| 703/1621 [13:46<14:01, 1.09it/s] 43%\|████▎ \| 704/1621 [13:47<13:39, 1.12it/s] 43%\|████▎ \| 705/1621 [13:48<14:14, 1.07it/s] 44%\|████�
	0: {'loss': 0.2935, 'grad_norm': 0.30645808053615586, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.44}
	0: �� \| 706/1621 [13:48<13:52, 1.10it/s] 44%\|████▎ \| 707/1621 [13:49<14:19, 1.06it/s] 44%\|████▎ \| 708/1621 [13:50<13:56, 1.09it/s] 44%\|████▎ \| 709/1621 [13:51<13:40, 1.11it/s] 44%\|████▍ \| 710/1621 [13:52<13:35, 1.12it/s] 44%\|████▍ \| 710/1621 [13:52<13:35, 1.12it/s] 44%\|████▍ \| 711/1621 [13:53<13:28, 1.13it/s] 44%\|████▍ \| 712/1621 [13:54<13:17, 1.14it/s] 44%\|████▍ \| 713/1621 [13:55<13:18, 1.14it/s] 44%\|████▍ \| 714/1621 [13:56<13:56, 1.08it/s] 44%\|████▍ \| 715/1621 [13:57<14:25, 1.05it/s] 44%\|████▍ \| 716/1621 [13:58<14:03, 1.07it/s] 44%\|████▍ \| 717/1621 [13:58<13:40, 1.10it/s] 44%\|████▍ \| 718/1621 [13:59<13:27, 1.12it/s] 44%\|████▍ \| 719/1621 [14:00<13:21, 1.13it/s] 44%\|████▍ \| 720/1621 [14:01<13:10, 1.14it/s]
	0: {'loss': 0.2951, 'grad_norm': 0.34070474789654825, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.44}
	0: {'loss': 0.3011, 'grad_norm': 0.3202334487749824, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.45}
	0: 44%\|████▍ \| 720/1621 [14:01<13:10, 1.14it/s] 44%\|████▍ \| 721/1621 [14:02<13:13, 1.13it/s] 45%\|████▍ \| 722/1621 [14:03<13:02, 1.15it/s] 45%\|████▍ \| 723/1621 [14:04<13:25, 1.12it/s] 45%\|████▍ \| 724/1621 [14:05<13:14, 1.13it/s] 45%\|████▍ \| 725/1621 [14:05<13:11, 1.13it/s] 45%\|████▍ \| 726/1621 [14:06<13:00, 1.15it/s] 45%\|████▍ \| 727/1621 [14:07<12:53, 1.16it/s] 45%\|████▍ \| 728/1621 [14:08<12:52, 1.16it/s] 45%\|████▍ \| 729/1621 [14:09<12:49, 1.16it/s] 45%\|████▌ \| 730/1621 [14:10<12:47, 1.16it/s] 45%\|████▌ \| 730/1621 [14:10<12:47, 1.16it/s] 45%\|████▌ \| 731/1621 [14:11<12:46, 1.16it/s] 45%\|████▌ \| 732/1621 [14:11<12:43, 1.16it/s] 45%\|████▌ \| 733/1621 [14:12<12:39, 1.17it/s] 45%\|████�
	0: {'loss': 0.2972, 'grad_norm': 0.3019359123796448, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.46}
	0: � \| 734/1621 [14:13<12:36, 1.17it/s] 45%\|████▌ \| 735/1621 [14:14<12:34, 1.17it/s] 45%\|████▌ \| 736/1621 [14:15<12:32, 1.18it/s] 45%\|████▌ \| 737/1621 [14:16<12:29, 1.18it/s] 46%\|████▌ \| 738/1621 [14:17<12:28, 1.18it/s] 46%\|████▌ \| 739/1621 [14:17<12:29, 1.18it/s] 46%\|████▌ \| 740/1621 [14:18<12:32, 1.17it/s] 46%\|████▌ \| 740/1621 [14:18<12:32, 1.17it/s] 46%\|████▌ \| 741/1621 [14:19<12:33, 1.17it/s] 46%\|████▌ \| 742/1621 [14:20<12:28, 1.17it/s] 46%\|████▌ \| 743/1621 [14:21<12:41, 1.15it/s] 46%\|████▌ \| 744/1621 [14:22<12:39, 1.15it/s] 46%\|████▌ \| 745/1621 [14:23<12:33, 1.16it/s] 46%\|████▌ \| 746/1621 [14:23<12:33, 1.16it/s] 46%\|████▌ \| 747/1621 [14:24<12:31, 1.16it/s] 46%\|████▌ \| 748/1621 [14:25<12:35, 1.16it/s] 46%\|███�
	0: {'loss': 0.2934, 'grad_norm': 0.29846589457849965, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.46}
	0: {'loss': 0.3048, 'grad_norm': 0.31089870103073103, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.47}
	0: ��▌ \| 749/1621 [14:26<12:32, 1.16it/s] 46%\|████▋ \| 750/1621 [14:27<12:29, 1.16it/s] 46%\|████▋ \| 750/1621 [14:27<12:29, 1.16it/s] 46%\|████▋ \| 751/1621 [14:28<12:37, 1.15it/s] 46%\|████▋ \| 752/1621 [14:29<12:29, 1.16it/s] 46%\|████▋ \| 753/1621 [14:29<12:24, 1.17it/s] 47%\|████▋ \| 754/1621 [14:30<12:22, 1.17it/s] 47%\|████▋ \| 755/1621 [14:31<12:19, 1.17it/s] 47%\|████▋ \| 756/1621 [14:32<12:15, 1.18it/s] 47%\|████▋ \| 757/1621 [14:33<12:12, 1.18it/s] 47%\|████▋ \| 758/1621 [14:34<12:11, 1.18it/s] 47%\|████▋ \| 759/1621 [14:35<12:35, 1.14it/s] 47%\|████▋ \| 760/1621 [14:35<12:32, 1.14it/s] 47%\|████▋ \| 760/1621 [14:35<12:32, 1.14it/s] 47%\|████▋ \| 761/1621 [14:36<12:26, 1.15it/s] 47%\|████▋
	0: {'loss': 0.2954, 'grad_norm': 0.3055451871683868, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.48}
	0: \| 762/1621 [14:37<12:19, 1.16it/s] 47%\|████▋ \| 763/1621 [14:38<12:25, 1.15it/s] 47%\|████▋ \| 764/1621 [14:39<12:37, 1.13it/s] 47%\|████▋ \| 765/1621 [14:40<13:43, 1.04it/s] 47%\|████▋ \| 766/1621 [14:41<13:13, 1.08it/s] 47%\|████▋ \| 767/1621 [14:42<12:51, 1.11it/s] 47%\|████▋ \| 768/1621 [14:43<12:36, 1.13it/s] 47%\|████▋ \| 769/1621 [14:44<12:26, 1.14it/s] 48%\|████▊ \| 770/1621 [14:44<12:24, 1.14it/s] 48%\|████▊ \| 770/1621 [14:44<12:24, 1.14it/s] 48%\|████▊ \| 771/1621 [14:45<12:21, 1.15it/s] 48%\|████▊ \| 772/1621 [14:46<12:16, 1.15it/s] 48%\|████▊ \| 773/1621 [14:47<12:12, 1.16it/s] 48%\|████▊ \| 774/1621 [14:48<12:15, 1.15it/s] 48%\|████▊ \| 775/1621 [14:49<12:09, 1.16it/s] 48%\|████▊ \| 776/1621 [14:50<12:05, 1.17it/s] 48%\|███�
	0: {'loss': 0.2953, 'grad_norm': 0.310251278561544, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.48}
	0: {'loss': 0.2877, 'grad_norm': 0.32857198895345724, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.49}
	0: �▊ \| 777/1621 [14:50<12:04, 1.17it/s] 48%\|████▊ \| 778/1621 [14:51<12:07, 1.16it/s] 48%\|████▊ \| 779/1621 [14:52<12:04, 1.16it/s] 48%\|████▊ \| 780/1621 [14:53<12:02, 1.16it/s] 48%\|████▊ \| 780/1621 [14:53<12:02, 1.16it/s] 48%\|████▊ \| 781/1621 [14:54<12:03, 1.16it/s] 48%\|████▊ \| 782/1621 [14:55<12:05, 1.16it/s] 48%\|████▊ \| 783/1621 [14:56<12:06, 1.15it/s] 48%\|████▊ \| 784/1621 [14:57<12:20, 1.13it/s] 48%\|████▊ \| 785/1621 [14:57<12:14, 1.14it/s] 48%\|████▊ \| 786/1621 [14:58<12:10, 1.14it/s] 49%\|████▊ \| 787/1621 [14:59<12:09, 1.14it/s] 49%\|████▊ \| 788/1621 [15:00<12:02, 1.15it/s] 49%\|████▊ \| 789/1621 [15:01<12:32, 1.11it/s] 49%\|████▊ \| 790/1621 [15:02<12:21, 1.12it/s] 49%\|████▊
	0: {'loss': 0.2924, 'grad_norm': 0.2994049813187105, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.49}
	0: \| 790/1621 [15:02<12:21, 1.12it/s] 49%\|████▉ \| 791/1621 [15:03<12:13, 1.13it/s] 49%\|████▉ \| 792/1621 [15:04<12:04, 1.14it/s] 49%\|████▉ \| 793/1621 [15:04<12:14, 1.13it/s] 49%\|████▉ \| 794/1621 [15:05<12:06, 1.14it/s] 49%\|████▉ \| 795/1621 [15:06<12:02, 1.14it/s] 49%\|████▉ \| 796/1621 [15:07<11:54, 1.16it/s] 49%\|████▉ \| 797/1621 [15:08<11:48, 1.16it/s] 49%\|████▉ \| 798/1621 [15:09<11:44, 1.17it/s] 49%\|████▉ \| 799/1621 [15:10<11:43, 1.17it/s] 49%\|████▉ \| 800/1621 [15:10<11:40, 1.17it/s] 49%\|████▉ \| 800/1621 [15:10<11:40, 1.17it/s] 49%\|████▉ \| 801/1621 [15:11<11:41, 1.17it/s] 49%\|████▉ \| 802/1621 [15:12<11:37, 1.17it/s] 50%\|████▉ \| 803/1621 [15:13<11:40, 1.17it/s] 50%\|████▉ \| 804/1621 [15:14<11:39, 1.17it/s] 50%\|████
	0: {'loss': 0.2911, 'grad_norm': 0.32458055128509955, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.5}
	0: ▉ \| 805/1621 [15:15<11:45, 1.16it/s] 50%\|████▉ \| 806/1621 [15:16<11:40, 1.16it/s] 50%\|████▉ \| 807/1621 [15:16<11:36, 1.17it/s] 50%\|████▉ \| 808/1621 [15:17<12:20, 1.10it/s] 50%\|████▉ \| 809/1621 [15:18<12:07, 1.12it/s] 50%\|████▉ \| 810/1621 [15:19<12:08, 1.11it/s] 50%\|████▉ \| 810/1621 [15:19<12:08, 1.11it/s] 50%\|█████ \| 811/1621 [15:20<12:00, 1.12it/s] 50%\|█████ \| 812/1621 [15:21<12:01, 1.12it/s] 50%\|█████ \| 813/1621 [15:22<11:51, 1.14it/s] 50%\|█████ \| 814/1621 [15:23<11:47, 1.14it/s] 50%\|█████ \| 815/1621 [15:24<11:40, 1.15it/s] 50%\|█████ \| 816/1621 [15:24<11:38, 1.15it/s] 50%\|█████ \| 817/1621 [15:25<11:34, 1.16it/s] 50%\|█████ \| 818/1621 [15:26<11:32, 1.16it/s] 51%\|█████ \| 819/1621 [15:27<11:34, 1.15it/s] 51%\|██�
	0: {'loss': 0.2926, 'grad_norm': 0.3005265468669908, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.51}
	0: {'loss': 0.3011, 'grad_norm': 0.3180393440770147, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.51}
	0: �██ \| 820/1621 [15:28<12:00, 1.11it/s] 51%\|█████ \| 820/1621 [15:28<12:00, 1.11it/s] 51%\|█████ \| 821/1621 [15:29<11:49, 1.13it/s] 51%\|█████ \| 822/1621 [15:30<11:39, 1.14it/s] 51%\|█████ \| 823/1621 [15:31<11:32, 1.15it/s] 51%\|█████ \| 824/1621 [15:32<11:43, 1.13it/s] 51%\|█████ \| 825/1621 [15:32<11:37, 1.14it/s] 51%\|█████ \| 826/1621 [15:33<11:30, 1.15it/s] 51%\|█████ \| 827/1621 [15:34<11:48, 1.12it/s] 51%\|█████ \| 828/1621 [15:35<11:46, 1.12it/s] 51%\|█████ \| 829/1621 [15:36<11:37, 1.14it/s] 51%\|█████ \| 830/1621 [15:37<11:32, 1.14it/s] 51%\|█████ \| 830/1621 [15:37<11:32, 1.14it/s] 51%\|█████▏ \| 831/1621 [15:38<11:50, 1.11it/s] 51%\|█████▏ \| 832/1621 [15:39<13:17, 1.01s/it] 51%\|███
	0: {'loss': 0.2884, 'grad_norm': 0.359753097877781, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.52}
	0: ██▏ \| 833/1621 [15:40<12:39, 1.04it/s] 51%\|█████▏ \| 834/1621 [15:41<12:10, 1.08it/s] 52%\|█████▏ \| 835/1621 [15:42<11:50, 1.11it/s] 52%\|█████▏ \| 836/1621 [15:42<11:51, 1.10it/s] 52%\|█████▏ \| 837/1621 [15:43<11:40, 1.12it/s] 52%\|█████▏ \| 838/1621 [15:44<12:19, 1.06it/s] 52%\|█████▏ \| 839/1621 [15:45<12:03, 1.08it/s] 52%\|█████▏ \| 840/1621 [15:46<11:58, 1.09it/s] 52%\|█████▏ \| 840/1621 [15:46<11:58, 1.09it/s] 52%\|█████▏ \| 841/1621 [15:47<11:48, 1.10it/s] 52%\|█████▏ \| 842/1621 [15:48<11:35, 1.12it/s] 52%\|█████▏ \| 843/1621 [15:49<11:53, 1.09it/s] 52%\|█████▏ \| 844/1621 [15:50<11:37, 1.11it/s] 52%\|█████▏ \| 845/1621 [15:51<11:26, 1.13it/s] 52%\|█████▏ \| 846/1621 [15:51<11:29, 1.12it/s] 52%\|█████▏ \| 847/1621 [15
	0: {'loss': 0.2858, 'grad_norm': 0.307823856703404, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.52}
	0: {'loss': 0.2923, 'grad_norm': 0.2996148986851413, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.53}
	0: :52<11:19, 1.14it/s] 52%\|█████▏ \| 848/1621 [15:53<11:27, 1.12it/s] 52%\|█████▏ \| 849/1621 [15:54<11:18, 1.14it/s] 52%\|█████▏ \| 850/1621 [15:55<11:10, 1.15it/s] 52%\|█████▏ \| 850/1621 [15:55<11:10, 1.15it/s] 52%\|█████▏ \| 851/1621 [15:56<11:23, 1.13it/s] 53%\|█████▎ \| 852/1621 [15:57<11:22, 1.13it/s] 53%\|█████▎ \| 853/1621 [15:58<11:14, 1.14it/s] 53%\|█████▎ \| 854/1621 [15:58<11:07, 1.15it/s] 53%\|█████▎ \| 855/1621 [15:59<11:03, 1.15it/s] 53%\|█████▎ \| 856/1621 [16:00<11:13, 1.14it/s] 53%\|█████▎ \| 857/1621 [16:01<11:35, 1.10it/s] 53%\|█████▎ \| 858/1621 [16:02<11:21, 1.12it/s] 53%\|█████▎ \| 859/1621 [16:03<11:13, 1.13it/s] 53%\|█████▎ \| 860/1621 [16:04<11:04, 1.14it/s] 53%\|███�
	0: {'loss': 0.2932, 'grad_norm': 0.2810728882724132, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.54}
	0: �█▎ \| 860/1621 [16:04<11:04, 1.14it/s] 53%\|█████▎ \| 861/1621 [16:05<11:34, 1.09it/s] 53%\|█████▎ \| 862/1621 [16:06<11:19, 1.12it/s] 53%\|█████▎ \| 863/1621 [16:07<11:12, 1.13it/s] 53%\|█████▎ \| 864/1621 [16:07<11:06, 1.14it/s] 53%\|█████▎ \| 865/1621 [16:08<10:58, 1.15it/s] 53%\|█████▎ \| 866/1621 [16:09<11:07, 1.13it/s] 53%\|█████▎ \| 867/1621 [16:10<10:58, 1.14it/s] 54%\|█████▎ \| 868/1621 [16:11<10:59, 1.14it/s] 54%\|█████▎ \| 869/1621 [16:12<10:52, 1.15it/s] 54%\|█████▎ \| 870/1621 [16:13<11:33, 1.08it/s] 54%\|█████▎ \| 870/1621 [16:13<11:33, 1.08it/s] 54%\|█████▎ \| 871/1621 [16:14<11:23, 1.10it/s] 54%\|█████▍ \| 872/1621 [16:15<11:20, 1.10it/s] 54%\|█████▍ \| 873/1621 [16:15<11:04, 1.12it/s] 54%\|█████▍ \| 874/1621 [16:1
	0: {'loss': 0.2937, 'grad_norm': 0.30282456722647566, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.54}
	0: 6<10:57, 1.14it/s] 54%\|█████▍ \| 875/1621 [16:17<10:56, 1.14it/s] 54%\|█████▍ \| 876/1621 [16:18<10:48, 1.15it/s] 54%\|█████▍ \| 877/1621 [16:19<10:44, 1.15it/s] 54%\|█████▍ \| 878/1621 [16:20<10:44, 1.15it/s] 54%\|█████▍ \| 879/1621 [16:21<10:40, 1.16it/s] 54%\|█████▍ \| 880/1621 [16:22<11:14, 1.10it/s] 54%\|█████▍ \| 880/1621 [16:22<11:14, 1.10it/s] 54%\|█████▍ \| 881/1621 [16:22<11:04, 1.11it/s] 54%\|█████▍ \| 882/1621 [16:23<10:57, 1.12it/s] 54%\|█████▍ \| 883/1621 [16:24<10:55, 1.13it/s] 55%\|█████▍ \| 884/1621 [16:25<10:49, 1.14it/s] 55%\|█████▍ \| 885/1621 [16:26<10:44, 1.14it/s] 55%\|█████▍ \| 886/1621 [16:27<10:37, 1.15it/s] 55%\|█████▍ \| 887/1621 [16:28<10:33, 1.16it/s] 55%\|█████▍ \| 888/1621 [16:29<10:34, 1.16it/s] 55%\|�
	0: {'loss': 0.2934, 'grad_norm': 0.3068512318705077, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.55}
	0: {'loss': 0.2955, 'grad_norm': 0.303815459875142, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.56}
	0: �████▍ \| 889/1621 [16:29<10:28, 1.16it/s] 55%\|█████▍ \| 890/1621 [16:30<10:28, 1.16it/s] 55%\|█████▍ \| 890/1621 [16:30<10:28, 1.16it/s] 55%\|█████▍ \| 891/1621 [16:31<10:36, 1.15it/s] 55%\|█████▌ \| 892/1621 [16:32<10:36, 1.15it/s] 55%\|█████▌ \| 893/1621 [16:33<10:43, 1.13it/s] 55%\|█████▌ \| 894/1621 [16:34<10:36, 1.14it/s] 55%\|█████▌ \| 895/1621 [16:35<10:29, 1.15it/s] 55%\|█████▌ \| 896/1621 [16:35<10:30, 1.15it/s] 55%\|█████▌ \| 897/1621 [16:36<10:31, 1.15it/s] 55%\|█████▌ \| 898/1621 [16:37<10:28, 1.15it/s] 55%\|█████▌ \| 899/1621 [16:38<10:25, 1.15it/s] 56%\|█████▌ \| 900/1621 [16:39<10:23, 1.16it/s] 56%\|█████▌ \| 900/1621 [16:39<10:23, 1.16it/s] 56%\|█████▌ \| 901/1621 [16:40<
	0: {'loss': 0.2879, 'grad_norm': 0.30668180739641193, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.56}
	0: 10:23, 1.16it/s] 56%\|█████▌ \| 902/1621 [16:41<10:56, 1.10it/s] 56%\|█████▌ \| 903/1621 [16:42<10:47, 1.11it/s] 56%\|█████▌ \| 904/1621 [16:43<10:41, 1.12it/s] 56%\|█████▌ \| 905/1621 [16:43<10:33, 1.13it/s] 56%\|█████▌ \| 906/1621 [16:44<10:32, 1.13it/s] 56%\|█████▌ \| 907/1621 [16:45<10:29, 1.13it/s] 56%\|█████▌ \| 908/1621 [16:46<10:22, 1.14it/s] 56%\|█████▌ \| 909/1621 [16:47<10:16, 1.15it/s] 56%\|█████▌ \| 910/1621 [16:48<10:10, 1.16it/s] 56%\|█████▌ \| 910/1621 [16:48<10:10, 1.16it/s] 56%\|█████▌ \| 911/1621 [16:49<10:08, 1.17it/s] 56%\|█████▋ \| 912/1621 [16:49<10:08, 1.17it/s] 56%\|█████▋ \| 913/1621 [16:50<10:11, 1.16it/s] 56%\|█████▋ \| 914/1621 [16:51<10:15, 1.15it/s] 56%\|█████▋ \| 915/1621 [16:52<10:50, 1.09it/s] 57%\|█�
	0: {'loss': 0.2898, 'grad_norm': 0.30259565260728083, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.57}
	0: ��███▋ \| 916/1621 [16:53<10:35, 1.11it/s] 57%\|█████▋ \| 917/1621 [16:54<10:25, 1.13it/s] 57%\|█████▋ \| 918/1621 [16:55<10:17, 1.14it/s] 57%\|█████▋ \| 919/1621 [16:56<10:18, 1.13it/s] 57%\|█████▋ \| 920/1621 [16:57<10:16, 1.14it/s] 57%\|█████▋ \| 920/1621 [16:57<10:16, 1.14it/s] 57%\|█████▋ \| 921/1621 [16:58<10:28, 1.11it/s] 57%\|█████▋ \| 922/1621 [16:58<10:17, 1.13it/s] 57%\|█████▋ \| 923/1621 [16:59<10:31, 1.10it/s] 57%\|█████▋ \| 924/1621 [17:00<10:50, 1.07it/s] 57%\|█████▋ \| 925/1621 [17:01<10:32, 1.10it/s] 57%\|█████▋ \| 926/1621 [17:02<10:20, 1.12it/s] 57%\|█████▋ \| 927/1621 [17:03<10:47, 1.07it/s] 57%\|█████▋ \| 928/1621 [17:04<10:32, 1.10it/s] 57%\|█████▋ \| 929/1621 [17:05<10:19, 1.12it/s] 57%\|█████▋ \| 930/162
	0: {'loss': 0.2892, 'grad_norm': 0.28494124741923377, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.57}
	0: {'loss': 0.2933, 'grad_norm': 0.3026563536305431, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.58}
	0: 1 [17:06<10:13, 1.13it/s] 57%\|█████▋ \| 930/1621 [17:06<10:13, 1.13it/s] 57%\|█████▋ \| 931/1621 [17:07<10:04, 1.14it/s] 57%\|█████▋ \| 932/1621 [17:07<09:57, 1.15it/s] 58%\|█████▊ \| 933/1621 [17:08<10:04, 1.14it/s] 58%\|█████▊ \| 934/1621 [17:09<09:58, 1.15it/s] 58%\|█████▊ \| 935/1621 [17:10<09:54, 1.15it/s] 58%\|█████▊ \| 936/1621 [17:11<10:04, 1.13it/s] 58%\|█████▊ \| 937/1621 [17:12<09:57, 1.14it/s] 58%\|█████▊ \| 938/1621 [17:13<10:29, 1.08it/s] 58%\|█████▊ \| 939/1621 [17:14<10:16, 1.11it/s] 58%\|█████▊ \| 940/1621 [17:15<10:06, 1.12it/s] 58%\|█████▊ \| 940/1621 [17:15<10:06, 1.12it/s] 58%\|█████▊ \| 941/1621 [17:15<10:03, 1.13it/s] 58%\|█████▊ \| 942/1621 [17:16<09:55, 1.14it/s] 58%\|██
	0: {'loss': 0.2949, 'grad_norm': 0.2955013797109043, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.59}
	0: ███▊ \| 943/1621 [17:17<09:50, 1.15it/s] 58%\|█████▊ \| 944/1621 [17:18<09:45, 1.16it/s] 58%\|█████▊ \| 945/1621 [17:19<09:41, 1.16it/s] 58%\|█████▊ \| 946/1621 [17:20<09:40, 1.16it/s] 58%\|█████▊ \| 947/1621 [17:21<09:38, 1.17it/s] 58%\|█████▊ \| 948/1621 [17:21<09:35, 1.17it/s] 59%\|█████▊ \| 949/1621 [17:22<09:31, 1.18it/s] 59%\|█████▊ \| 950/1621 [17:23<09:30, 1.18it/s] 59%\|█████▊ \| 950/1621 [17:23<09:30, 1.18it/s] 59%\|█████▊ \| 951/1621 [17:24<09:46, 1.14it/s] 59%\|█████▊ \| 952/1621 [17:25<09:44, 1.15it/s] 59%\|█████▉ \| 953/1621 [17:26<09:40, 1.15it/s] 59%\|█████▉ \| 954/1621 [17:27<09:44, 1.14it/s] 59%\|█████▉ \| 955/1621 [17:28<09:53, 1.12it/s] 59%\|█████▉ \| 956/1621 [17:28<09:45, 1.14it/s] 59%\|█████▉ \| 957/1621
	0: {'loss': 0.2868, 'grad_norm': 0.28031507522996657, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.59}
	0: {'loss': 0.2978, 'grad_norm': 0.29965929088247534, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.6}
	0: [17:29<09:45, 1.13it/s] 59%\|█████▉ \| 958/1621 [17:30<09:38, 1.15it/s] 59%\|█████▉ \| 959/1621 [17:31<09:37, 1.15it/s] 59%\|█████▉ \| 960/1621 [17:32<09:33, 1.15it/s] 59%\|█████▉ \| 960/1621 [17:32<09:33, 1.15it/s] 59%\|█████▉ \| 961/1621 [17:33<09:35, 1.15it/s] 59%\|█████▉ \| 962/1621 [17:34<09:30, 1.15it/s] 59%\|█████▉ \| 963/1621 [17:34<09:27, 1.16it/s] 59%\|█████▉ \| 964/1621 [17:35<09:23, 1.17it/s] 60%\|█████▉ \| 965/1621 [17:36<09:23, 1.16it/s] 60%\|█████▉ \| 966/1621 [17:37<09:19, 1.17it/s] 60%\|█████▉ \| 967/1621 [17:38<09:19, 1.17it/s] 60%\|█████▉ \| 968/1621 [17:39<09:20, 1.17it/s] 60%\|█████▉ \| 969/1621 [17:40<09:20, 1.16it/s] 60%\|█████▉ \| 970/1621 [17:40<09:20, 1.16it/s] 60%\|██�
	0: [2025-09-02 18:53:55,696] [INFO] [axolotl.core.trainers.base._save:613] [PID:801836] [RANK:0] Saving model checkpoint to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0/checkpoint-975[39m
	0: [2025-09-02 18:53:58,194] [INFO] [axolotl.core.trainers.base._save:662] [PID:801836] [RANK:0] Saving Trainer.data_collator.tokenizer by default as Trainer.processing_class is `None`[39m
	0: {'loss': 0.2943, 'grad_norm': 0.29624852844606697, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.6}
	0: �██▉ \| 970/1621 [17:40<09:20, 1.16it/s] 60%\|█████▉ \| 971/1621 [17:41<09:18, 1.16it/s] 60%\|█████▉ \| 972/1621 [17:42<09:15, 1.17it/s] 60%\|██████ \| 973/1621 [17:43<09:13, 1.17it/s] 60%\|██████ \| 974/1621 [17:44<09:13, 1.17it/s] 60%\|██████ \| 975/1621 [17:45<09:21, 1.15it/s] 60%\|██████ \| 976/1621 [17:50<23:43, 2.21s/it] 60%\|██████ \| 977/1621 [17:51<19:19, 1.80s/it] 60%\|██████ \| 978/1621 [17:52<16:13, 1.51s/it] 60%\|██████ \| 979/1621 [17:53<14:13, 1.33s/it] 60%\|██████ \| 980/1621 [17:54<12:43, 1.19s/it] 60%\|██████ \| 980/1621 [17:54<12:43, 1.19s/it] 61%\|██████ \| 981/1621 [17:54<11:42, 1.10s/it] 61%\|██████ \| 982/1621 [17:55<10:54, 1.02s/it] 61%\|██████ \| 983/1621 [17:56<10:22, 1.02it/s] 61%\|██████ \| 984/1621 [1
	0: {'loss': 0.296, 'grad_norm': 0.30350730138638954, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.61}
	0: 7:57<10:17, 1.03it/s] 61%\|██████ \| 985/1621 [17:58<09:54, 1.07it/s] 61%\|██████ \| 986/1621 [17:59<09:37, 1.10it/s] 61%\|██████ \| 987/1621 [18:00<09:34, 1.10it/s] 61%\|██████ \| 988/1621 [18:01<09:27, 1.12it/s] 61%\|██████ \| 989/1621 [18:01<09:16, 1.13it/s] 61%\|██████ \| 990/1621 [18:02<09:13, 1.14it/s] 61%\|██████ \| 990/1621 [18:02<09:13, 1.14it/s] 61%\|██████ \| 991/1621 [18:03<09:08, 1.15it/s] 61%\|██████ \| 992/1621 [18:04<09:04, 1.15it/s] 61%\|██████▏ \| 993/1621 [18:05<09:00, 1.16it/s] 61%\|██████▏ \| 994/1621 [18:06<08:57, 1.17it/s] 61%\|██████▏ \| 995/1621 [18:07<08:54, 1.17it/s] 61%\|██████▏ \| 996/1621 [18:08<09:29, 1.10it/s] 62%\|██████▏ \| 997/1621 [18:08<09:18, 1.12it/s] 62%\|██████▏ \| 998/1621 [18:09<09:08, 1.
	0: {'loss': 0.2983, 'grad_norm': 0.3203748629162452, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.62}
	0: {'loss': 0.2842, 'grad_norm': 0.2997338003534859, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.62}
	0: 14it/s] 62%\|██████▏ \| 999/1621 [18:10<09:02, 1.15it/s] 62%\|██████▏ \| 1000/1621 [18:11<09:02, 1.14it/s] 62%\|██████▏ \| 1000/1621 [18:11<09:02, 1.14it/s] 62%\|██████▏ \| 1001/1621 [18:12<08:59, 1.15it/s] 62%\|██████▏ \| 1002/1621 [18:13<08:58, 1.15it/s] 62%\|██████▏ \| 1003/1621 [18:14<08:56, 1.15it/s] 62%\|██████▏ \| 1004/1621 [18:14<08:53, 1.16it/s] 62%\|██████▏ \| 1005/1621 [18:15<08:52, 1.16it/s] 62%\|██████▏ \| 1006/1621 [18:16<08:50, 1.16it/s] 62%\|██████▏ \| 1007/1621 [18:17<08:46, 1.17it/s] 62%\|██████▏ \| 1008/1621 [18:18<08:43, 1.17it/s] 62%\|██████▏ \| 1009/1621 [18:19<09:01, 1.13it/s] 62%\|██████▏ \| 1010/1621 [18:20<08:56, 1.14it/s] 62%\|██████▏ \| 1010/1621 [18:20<08:56
	0: {'loss': 0.2849, 'grad_norm': 0.31338661689855674, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.63}
	0: , 1.14it/s] 62%\|██████▏ \| 1011/1621 [18:21<08:51, 1.15it/s] 62%\|██████▏ \| 1012/1621 [18:22<09:07, 1.11it/s] 62%\|██████▏ \| 1013/1621 [18:22<09:13, 1.10it/s] 63%\|██████▎ \| 1014/1621 [18:23<09:03, 1.12it/s] 63%\|██████▎ \| 1015/1621 [18:24<08:56, 1.13it/s] 63%\|██████▎ \| 1016/1621 [18:25<08:59, 1.12it/s] 63%\|██████▎ \| 1017/1621 [18:26<08:58, 1.12it/s] 63%\|██████▎ \| 1018/1621 [18:27<09:01, 1.11it/s] 63%\|██████▎ \| 1019/1621 [18:28<08:53, 1.13it/s] 63%\|██████▎ \| 1020/1621 [18:29<08:49, 1.14it/s] 63%\|██████▎ \| 1020/1621 [18:29<08:49, 1.14it/s] 63%\|██████▎ \| 1021/1621 [18:30<08:52, 1.13it/s] 63%\|██████▎ \| 1022/1621 [18:30<08:52, 1.13it/s] 63%\|██████▎ \| 1023/1621 [18:31<08:44, 1.14it/s] 63%\|██████▎ \| 102
	0: {'loss': 0.2896, 'grad_norm': 0.317479701582765, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.64}
	0: 4/1621 [18:32<08:39, 1.15it/s] 63%\|██████▎ \| 1025/1621 [18:33<08:45, 1.13it/s] 63%\|██████▎ \| 1026/1621 [18:34<09:14, 1.07it/s] 63%\|██████▎ \| 1027/1621 [18:35<08:59, 1.10it/s] 63%\|██████▎ \| 1028/1621 [18:36<09:01, 1.09it/s] 63%\|██████▎ \| 1029/1621 [18:37<08:53, 1.11it/s] 64%\|██████▎ \| 1030/1621 [18:38<08:56, 1.10it/s] 64%\|██████▎ \| 1030/1621 [18:38<08:56, 1.10it/s] 64%\|██████▎ \| 1031/1621 [18:39<09:00, 1.09it/s] 64%\|██████▎ \| 1032/1621 [18:40<08:57, 1.10it/s] 64%\|██████▎ \| 1033/1621 [18:40<08:46, 1.12it/s] 64%\|██████▍ \| 1034/1621 [18:41<08:39, 1.13it/s] 64%\|██████▍ \| 1035/1621 [18:42<08:34, 1.14it/s] 64%\|██████▍ \| 1036/1621 [18:43<08:40, 1.12it/s] 64%\|██████▍ \| 1037/1621 [18:44<08:35, 1.13it/s] 64%\|███�
	0: {'loss': 0.2973, 'grad_norm': 0.2949775076536377, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.64}
	0: {'loss': 0.2888, 'grad_norm': 0.3078098551918824, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.65}
	0: ��██▍ \| 1038/1621 [18:45<08:30, 1.14it/s] 64%\|██████▍ \| 1039/1621 [18:46<08:44, 1.11it/s] 64%\|██████▍ \| 1040/1621 [18:47<08:40, 1.12it/s] 64%\|██████▍ \| 1040/1621 [18:47<08:40, 1.12it/s] 64%\|██████▍ \| 1041/1621 [18:48<09:07, 1.06it/s] 64%\|██████▍ \| 1042/1621 [18:48<08:51, 1.09it/s] 64%\|██████▍ \| 1043/1621 [18:49<08:37, 1.12it/s] 64%\|██████▍ \| 1044/1621 [18:50<08:31, 1.13it/s] 64%\|██████▍ \| 1045/1621 [18:51<08:24, 1.14it/s] 65%\|██████▍ \| 1046/1621 [18:52<08:20, 1.15it/s] 65%\|██████▍ \| 1047/1621 [18:53<08:16, 1.16it/s] 65%\|██████▍ \| 1048/1621 [18:54<08:14, 1.16it/s] 65%\|██████▍ \| 1049/1621 [18:54<08:15, 1.16it/s] 65%\|██████▍ \| 1050/1621 [18:55<08:12, 1.16it/s] 65%\|█�
	0: {'loss': 0.2936, 'grad_norm': 0.3170740850922587, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.65}
	0: ��████▍ \| 1050/1621 [18:55<08:12, 1.16it/s] 65%\|██████▍ \| 1051/1621 [18:56<08:24, 1.13it/s] 65%\|██████▍ \| 1052/1621 [18:57<08:44, 1.08it/s] 65%\|██████▍ \| 1053/1621 [18:58<08:35, 1.10it/s] 65%\|██████▌ \| 1054/1621 [18:59<08:23, 1.13it/s] 65%\|██████▌ \| 1055/1621 [19:00<08:16, 1.14it/s] 65%\|██████▌ \| 1056/1621 [19:01<08:12, 1.15it/s] 65%\|██████▌ \| 1057/1621 [19:02<08:08, 1.15it/s] 65%\|██████▌ \| 1058/1621 [19:02<08:09, 1.15it/s] 65%\|██████▌ \| 1059/1621 [19:03<08:17, 1.13it/s] 65%\|██████▌ \| 1060/1621 [19:04<08:22, 1.12it/s] 65%\|██████▌ \| 1060/1621 [19:04<08:22, 1.12it/s] 65%\|██████▌ \| 1061/1621 [19:05<08:14, 1.13it/s] 66%\|██████▌ \| 1062/1621 [19:06<08:09, 1.14it/s] 66%\|██████▌ \| 1063/1621 [19:07<08:15,
	0: {'loss': 0.2845, 'grad_norm': 0.3117559688954944, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.66}
	0: 1.13it/s] 66%\|██████▌ \| 1064/1621 [19:08<08:11, 1.13it/s] 66%\|██████▌ \| 1065/1621 [19:09<08:04, 1.15it/s] 66%\|██████▌ \| 1066/1621 [19:10<08:32, 1.08it/s] 66%\|██████▌ \| 1067/1621 [19:11<08:20, 1.11it/s] 66%\|██████▌ \| 1068/1621 [19:11<08:15, 1.12it/s] 66%\|██████▌ \| 1069/1621 [19:12<08:15, 1.11it/s] 66%\|██████▌ \| 1070/1621 [19:13<08:11, 1.12it/s] 66%\|██████▌ \| 1070/1621 [19:13<08:11, 1.12it/s] 66%\|██████▌ \| 1071/1621 [19:14<08:05, 1.13it/s] 66%\|██████▌ \| 1072/1621 [19:15<08:06, 1.13it/s] 66%\|██████▌ \| 1073/1621 [19:16<08:41, 1.05it/s] 66%\|██████▋ \| 1074/1621 [19:17<08:25, 1.08it/s] 66%\|██████▋ \| 1075/1621 [19:18<08:18, 1.10it/s] 66%\|██████▋ \| 1076/1621 [19:19<08:11, 1.11it/s] 66%\|██████▋ \| 1077/1
	0: {'loss': 0.2835, 'grad_norm': 0.29256036199346086, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.67}
	0: 621 [19:20<08:05, 1.12it/s] 67%\|██████▋ \| 1078/1621 [19:21<08:20, 1.09it/s] 67%\|██████▋ \| 1079/1621 [19:21<08:08, 1.11it/s] 67%\|██████▋ \| 1080/1621 [19:22<07:59, 1.13it/s] 67%\|██████▋ \| 1080/1621 [19:22<07:59, 1.13it/s] 67%\|██████▋ \| 1081/1621 [19:23<07:52, 1.14it/s] 67%\|██████▋ \| 1082/1621 [19:24<07:51, 1.14it/s] 67%\|██████▋ \| 1083/1621 [19:25<08:01, 1.12it/s] 67%\|██████▋ \| 1084/1621 [19:26<07:54, 1.13it/s] 67%\|██████▋ \| 1085/1621 [19:27<07:47, 1.15it/s] 67%\|██████▋ \| 1086/1621 [19:27<07:46, 1.15it/s] 67%\|██████▋ \| 1087/1621 [19:28<07:58, 1.12it/s] 67%\|██████▋ \| 1088/1621 [19:29<07:55, 1.12it/s] 67%\|██████▋ \| 1089/1621 [19:30<07:50, 1.13it/s] 67%\|██████▋ \| 1090/1621 [19:31<07:53, 1.12it/s]
	0: {'loss': 0.2824, 'grad_norm': 0.29072721767613235, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.67}
	0: {'loss': 0.2923, 'grad_norm': 0.29738132555989805, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.68}
	0: 67%\|██████▋ \| 1090/1621 [19:31<07:53, 1.12it/s] 67%\|██████▋ \| 1091/1621 [19:32<07:45, 1.14it/s] 67%\|██████▋ \| 1092/1621 [19:33<07:42, 1.14it/s] 67%\|██████▋ \| 1093/1621 [19:34<07:42, 1.14it/s] 67%\|██████▋ \| 1094/1621 [19:35<07:38, 1.15it/s] 68%\|██████▊ \| 1095/1621 [19:35<07:36, 1.15it/s] 68%\|██████▊ \| 1096/1621 [19:36<07:40, 1.14it/s] 68%\|██████▊ \| 1097/1621 [19:37<07:35, 1.15it/s] 68%\|██████▊ \| 1098/1621 [19:38<07:40, 1.14it/s] 68%\|██████▊ \| 1099/1621 [19:39<07:37, 1.14it/s] 68%\|██████▊ \| 1100/1621 [19:40<07:36, 1.14it/s] 68%\|██████▊ \| 1100/1621 [19:40<07:36, 1.14it/s] 68%\|██████▊ \| 1101/1621 [19:41<07:33, 1.15it/s] 68%\|██████▊ \| 1102/1621 [19:41<07:28, 1.16it/s] 68%\|██�
	0: {'loss': 0.2888, 'grad_norm': 0.39509367378956034, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.68}
	0: ��███▊ \| 1103/1621 [19:42<07:26, 1.16it/s] 68%\|██████▊ \| 1104/1621 [19:43<07:27, 1.16it/s] 68%\|██████▊ \| 1105/1621 [19:44<07:23, 1.16it/s] 68%\|██████▊ \| 1106/1621 [19:45<07:57, 1.08it/s] 68%\|██████▊ \| 1107/1621 [19:46<07:49, 1.10it/s] 68%\|██████▊ \| 1108/1621 [19:47<07:37, 1.12it/s] 68%\|██████▊ \| 1109/1621 [19:48<07:35, 1.12it/s] 68%\|██████▊ \| 1110/1621 [19:49<07:36, 1.12it/s] 68%\|██████▊ \| 1110/1621 [19:49<07:36, 1.12it/s] 69%\|██████▊ \| 1111/1621 [19:50<07:32, 1.13it/s] 69%\|██████▊ \| 1112/1621 [19:51<07:58, 1.06it/s] 69%\|██████▊ \| 1113/1621 [19:51<07:47, 1.09it/s] 69%\|██████▊ \| 1114/1621 [19:52<07:37, 1.11it/s] 69%\|██████▉ \| 1115/1621 [19:53<07:28, 1.13it/s] 69%\|██████▉ \| 1116/1621 [19:54<07:24, 1.1
	0: {'loss': 0.292, 'grad_norm': 0.30996483704428673, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.69}
	0: 4it/s] 69%\|██████▉ \| 1117/1621 [19:55<07:45, 1.08it/s] 69%\|██████▉ \| 1118/1621 [19:56<07:33, 1.11it/s] 69%\|██████▉ \| 1119/1621 [19:57<07:23, 1.13it/s] 69%\|██████▉ \| 1120/1621 [19:58<07:17, 1.15it/s] 69%\|██████▉ \| 1120/1621 [19:58<07:17, 1.15it/s] 69%\|██████▉ \| 1121/1621 [19:59<07:19, 1.14it/s] 69%\|██████▉ \| 1122/1621 [19:59<07:26, 1.12it/s] 69%\|██████▉ \| 1123/1621 [20:00<07:18, 1.13it/s] 69%\|██████▉ \| 1124/1621 [20:01<07:15, 1.14it/s] 69%\|██████▉ \| 1125/1621 [20:02<07:44, 1.07it/s] 69%\|██████▉ \| 1126/1621 [20:03<07:31, 1.10it/s] 70%\|██████▉ \| 1127/1621 [20:04<07:22, 1.12it/s] 70%\|██████▉ \| 1128/1621 [20:05<07:26, 1.10it/s] 70%\|██████▉ \| 1129/1621 [20:06<07:31, 1.09it/s] 70%\|██████▉ \| 1130/1621
	0: {'loss': 0.2888, 'grad_norm': 0.3063733535961954, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.7}
	0: {'loss': 0.2922, 'grad_norm': 0.3416852250077931, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.7}
	0: [20:07<07:24, 1.11it/s] 70%\|██████▉ \| 1130/1621 [20:07<07:24, 1.11it/s] 70%\|██████▉ \| 1131/1621 [20:08<07:16, 1.12it/s] 70%\|██████▉ \| 1132/1621 [20:08<07:12, 1.13it/s] 70%\|██████▉ \| 1133/1621 [20:09<07:07, 1.14it/s] 70%\|██████▉ \| 1134/1621 [20:10<07:03, 1.15it/s] 70%\|███████ \| 1135/1621 [20:11<07:45, 1.04it/s] 70%\|███████ \| 1136/1621 [20:12<07:29, 1.08it/s] 70%\|███████ \| 1137/1621 [20:13<07:26, 1.08it/s] 70%\|███████ \| 1138/1621 [20:14<07:15, 1.11it/s] 70%\|███████ \| 1139/1621 [20:15<07:34, 1.06it/s] 70%\|███████ \| 1140/1621 [20:16<07:31, 1.06it/s] 70%\|███████ \| 1140/1621 [20:16<07:31, 1.06it/s] 70%\|███████ \| 1141/1621 [20:17<07:36, 1.05it/s] 70%\|███████ \| 114
	0: {'loss': 0.2925, 'grad_norm': 0.30884469264741793, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.71}
	0: 2/1621 [20:18<07:27, 1.07it/s] 71%\|███████ \| 1143/1621 [20:19<07:15, 1.10it/s] 71%\|███████ \| 1144/1621 [20:19<07:07, 1.12it/s] 71%\|███████ \| 1145/1621 [20:20<07:00, 1.13it/s] 71%\|███████ \| 1146/1621 [20:21<06:54, 1.15it/s] 71%\|███████ \| 1147/1621 [20:22<06:51, 1.15it/s] 71%\|███████ \| 1148/1621 [20:23<06:54, 1.14it/s] 71%\|███████ \| 1149/1621 [20:24<06:49, 1.15it/s] 71%\|███████ \| 1150/1621 [20:25<06:47, 1.16it/s] 71%\|███████ \| 1150/1621 [20:25<06:47, 1.16it/s] 71%\|███████ \| 1151/1621 [20:25<06:45, 1.16it/s] 71%\|███████ \| 1152/1621 [20:26<06:53, 1.13it/s] 71%\|███████ \| 1153/1621 [20:27<06:50, 1.14it/s] 71%\|███████ \| 1154/1621 [20:28<06:45, 1.15it/s] 71%\|███████▏ \| 1155/1621 [20:29<06:43, 1.15it/s] 71%\|██�
	0: {'loss': 0.2894, 'grad_norm': 0.3044894747315433, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.72}
	0: �████▏ \| 1156/1621 [20:30<06:43, 1.15it/s] 71%\|███████▏ \| 1157/1621 [20:31<06:39, 1.16it/s] 71%\|███████▏ \| 1158/1621 [20:32<06:37, 1.16it/s] 71%\|███████▏ \| 1159/1621 [20:32<06:38, 1.16it/s] 72%\|███████▏ \| 1160/1621 [20:33<06:51, 1.12it/s] 72%\|███████▏ \| 1160/1621 [20:33<06:51, 1.12it/s] 72%\|███████▏ \| 1161/1621 [20:34<06:46, 1.13it/s] 72%\|███████▏ \| 1162/1621 [20:35<06:39, 1.15it/s] 72%\|███████▏ \| 1163/1621 [20:36<06:41, 1.14it/s] 72%\|███████▏ \| 1164/1621 [20:37<06:40, 1.14it/s] 72%\|███████▏ \| 1165/1621 [20:38<06:38, 1.14it/s] 72%\|███████▏ \| 1166/1621 [20:39<06:34, 1.15it/s] 72%\|███████▏ \| 1167/1621 [20:39<06:31, 1.16it/s] 72%\|███████▏ \| 1168/1621 [20:40<06:28, 1.17it/s] 72%\|███████▏ \|
	0: {'loss': 0.2951, 'grad_norm': 0.30433585143251524, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.72}
	0: {'loss': 0.2873, 'grad_norm': 0.2914456090348807, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.73}
	0: 1169/1621 [20:41<06:35, 1.14it/s] 72%\|███████▏ \| 1170/1621 [20:42<06:31, 1.15it/s] 72%\|███████▏ \| 1170/1621 [20:42<06:31, 1.15it/s] 72%\|███████▏ \| 1171/1621 [20:43<06:32, 1.15it/s] 72%\|███████▏ \| 1172/1621 [20:44<06:35, 1.14it/s] 72%\|███████▏ \| 1173/1621 [20:45<06:30, 1.15it/s] 72%\|███████▏ \| 1174/1621 [20:46<06:26, 1.16it/s] 72%\|███████▏ \| 1175/1621 [20:46<06:22, 1.17it/s] 73%\|███████▎ \| 1176/1621 [20:47<06:21, 1.17it/s] 73%\|███████▎ \| 1177/1621 [20:48<06:23, 1.16it/s] 73%\|███████▎ \| 1178/1621 [20:49<06:22, 1.16it/s] 73%\|███████▎ \| 1179/1621 [20:50<06:25, 1.15it/s] 73%\|███████▎ \| 1180/1621 [20:51<06:22, 1.15it/s] 73%\|███████▎ \| 1180/1621 [20:51<06:22, 1.15it/s
	0: {'loss': 0.2904, 'grad_norm': 0.29539333636292997, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.73}
	0: ] 73%\|███████▎ \| 1181/1621 [20:52<06:20, 1.16it/s] 73%\|███████▎ \| 1182/1621 [20:52<06:20, 1.15it/s] 73%\|███████▎ \| 1183/1621 [20:53<06:17, 1.16it/s] 73%\|███████▎ \| 1184/1621 [20:54<06:15, 1.17it/s] 73%\|███████▎ \| 1185/1621 [20:55<06:13, 1.17it/s] 73%\|███████▎ \| 1186/1621 [20:56<06:12, 1.17it/s] 73%\|███████▎ \| 1187/1621 [20:57<06:10, 1.17it/s] 73%\|███████▎ \| 1188/1621 [20:58<06:10, 1.17it/s] 73%\|███████▎ \| 1189/1621 [20:58<06:09, 1.17it/s] 73%\|███████▎ \| 1190/1621 [20:59<06:08, 1.17it/s] 73%\|███████▎ \| 1190/1621 [20:59<06:08, 1.17it/s] 73%\|███████▎ \| 1191/1621 [21:00<06:09, 1.16it/s] 74%\|███████▎ \| 1192/1621 [21:01<06:08, 1.17it/s] 74%\|███████▎ \| 1193/1621 [21:02<06:08, 1.16it/s] 74%\|████
	0: {'loss': 0.2846, 'grad_norm': 0.29449247766474873, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.74}
	0: ███▎ \| 1194/1621 [21:03<06:07, 1.16it/s] 74%\|███████▎ \| 1195/1621 [21:04<06:20, 1.12it/s] 74%\|███████▍ \| 1196/1621 [21:05<06:15, 1.13it/s] 74%\|███████▍ \| 1197/1621 [21:05<06:11, 1.14it/s] 74%\|███████▍ \| 1198/1621 [21:06<06:21, 1.11it/s] 74%\|███████▍ \| 1199/1621 [21:07<06:20, 1.11it/s] 74%\|███████▍ \| 1200/1621 [21:08<06:13, 1.13it/s] 74%\|███████▍ \| 1200/1621 [21:08<06:13, 1.13it/s] 74%\|███████▍ \| 1201/1621 [21:09<06:21, 1.10it/s] 74%\|███████▍ \| 1202/1621 [21:10<06:14, 1.12it/s] 74%\|███████▍ \| 1203/1621 [21:11<06:11, 1.13it/s] 74%\|███████▍ \| 1204/1621 [21:12<06:05, 1.14it/s] 74%\|███████▍ \| 1205/1621 [21:13<06:02, 1.15it/s] 74%\|███████▍ \| 1206/1621 [21:14<06:26, 1.07it/s] 74%\|███████▍ \| 120
	0: {'loss': 0.2885, 'grad_norm': 0.2826708578899858, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.75}
	0: 7/1621 [21:14<06:14, 1.11it/s] 75%\|███████▍ \| 1208/1621 [21:15<06:09, 1.12it/s] 75%\|███████▍ \| 1209/1621 [21:16<06:03, 1.13it/s] 75%\|███████▍ \| 1210/1621 [21:17<06:00, 1.14it/s] 75%\|███████▍ \| 1210/1621 [21:17<06:00, 1.14it/s] 75%\|███████▍ \| 1211/1621 [21:18<06:03, 1.13it/s] 75%\|███████▍ \| 1212/1621 [21:19<06:00, 1.13it/s] 75%\|███████▍ \| 1213/1621 [21:20<05:56, 1.14it/s] 75%\|███████▍ \| 1214/1621 [21:21<05:54, 1.15it/s] 75%\|███████▍ \| 1215/1621 [21:21<06:00, 1.13it/s] 75%\|███████▌ \| 1216/1621 [21:22<05:57, 1.13it/s] 75%\|███████▌ \| 1217/1621 [21:23<05:54, 1.14it/s] 75%\|███████▌ \| 1218/1621 [21:24<05:50, 1.15it/s] 75%\|███████▌ \| 1219/1621 [21:25<05:47, 1.16it/s] 75%\|███████▌ \| 1220/1621 [21:26<05:48
	0: {'loss': 0.2881, 'grad_norm': 0.29371514774816154, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.75}
	0: {'loss': 0.2925, 'grad_norm': 0.3014384531364355, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.76}
	0: , 1.15it/s] 75%\|███████▌ \| 1220/1621 [21:26<05:48, 1.15it/s] 75%\|███████▌ \| 1221/1621 [21:27<05:45, 1.16it/s] 75%\|███████▌ \| 1222/1621 [21:27<05:42, 1.16it/s] 75%\|███████▌ \| 1223/1621 [21:28<05:46, 1.15it/s] 76%\|███████▌ \| 1224/1621 [21:29<05:44, 1.15it/s] 76%\|███████▌ \| 1225/1621 [21:30<05:41, 1.16it/s] 76%\|███████▌ \| 1226/1621 [21:31<05:38, 1.17it/s] 76%\|███████▌ \| 1227/1621 [21:32<05:37, 1.17it/s] 76%\|███████▌ \| 1228/1621 [21:33<05:38, 1.16it/s] 76%\|███████▌ \| 1229/1621 [21:34<05:36, 1.16it/s] 76%\|███████▌ \| 1230/1621 [21:34<05:33, 1.17it/s] 76%\|███████▌ \| 1230/1621 [21:34<05:33, 1.17it/s] 76%\|███████▌ \| 1231/1621 [21:35<05:32, 1.17it/s] 76%\|█████�
	0: {'loss': 0.2906, 'grad_norm': 0.3072217593018881, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.76}
	0: ��█▌ \| 1232/1621 [21:36<05:37, 1.15it/s] 76%\|███████▌ \| 1233/1621 [21:37<05:35, 1.16it/s] 76%\|███████▌ \| 1234/1621 [21:38<05:34, 1.16it/s] 76%\|███████▌ \| 1235/1621 [21:39<05:33, 1.16it/s] 76%\|███████▌ \| 1236/1621 [21:40<05:32, 1.16it/s] 76%\|███████▋ \| 1237/1621 [21:40<05:29, 1.16it/s] 76%\|███████▋ \| 1238/1621 [21:41<05:55, 1.08it/s] 76%\|███████▋ \| 1239/1621 [21:42<05:47, 1.10it/s] 76%\|███████▋ \| 1240/1621 [21:43<05:43, 1.11it/s] 76%\|███████▋ \| 1240/1621 [21:43<05:43, 1.11it/s] 77%\|███████▋ \| 1241/1621 [21:44<05:39, 1.12it/s] 77%\|███████▋ \| 1242/1621 [21:45<05:33, 1.14it/s] 77%\|███████▋ \| 1243/1621 [21:46<05:31, 1.14it/s] 77%\|███████▋ \| 1244/1621 [21:47<05:27, 1.15it/s] 77%\|███████▋ \| 1245/16
	0: {'loss': 0.284, 'grad_norm': 0.3038905157918577, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.77}
	0: 21 [21:48<05:34, 1.12it/s] 77%\|███████▋ \| 1246/1621 [21:49<05:44, 1.09it/s] 77%\|███████▋ \| 1247/1621 [21:49<05:35, 1.11it/s] 77%\|███████▋ \| 1248/1621 [21:50<05:29, 1.13it/s] 77%\|███████▋ \| 1249/1621 [21:51<05:29, 1.13it/s] 77%\|███████▋ \| 1250/1621 [21:52<05:25, 1.14it/s] 77%\|███████▋ \| 1250/1621 [21:52<05:25, 1.14it/s] 77%\|███████▋ \| 1251/1621 [21:53<05:21, 1.15it/s] 77%\|███████▋ \| 1252/1621 [21:54<05:18, 1.16it/s] 77%\|███████▋ \| 1253/1621 [21:55<05:16, 1.16it/s] 77%\|███████▋ \| 1254/1621 [21:55<05:16, 1.16it/s] 77%\|███████▋ \| 1255/1621 [21:56<05:14, 1.17it/s] 77%\|███████▋ \| 1256/1621 [21:57<05:14, 1.16it/s] 78%\|███████▊ \| 1257/1621 [21:58<05:12, 1.17it/s] 78%\|███████▊ \| 1258/1621 [21:59<05:09, 1
	0: {'loss': 0.29, 'grad_norm': 0.3050525690945563, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.78}
	0: {'loss': 0.2953, 'grad_norm': 0.31129474618367203, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.78}
	0: .17it/s] 78%\|███████▊ \| 1259/1621 [22:00<05:10, 1.16it/s] 78%\|███████▊ \| 1260/1621 [22:01<05:10, 1.16it/s] 78%\|███████▊ \| 1260/1621 [22:01<05:10, 1.16it/s] 78%\|███████▊ \| 1261/1621 [22:02<05:14, 1.14it/s] 78%\|███████▊ \| 1262/1621 [22:02<05:11, 1.15it/s] 78%\|███████▊ \| 1263/1621 [22:03<05:12, 1.14it/s] 78%\|███████▊ \| 1264/1621 [22:04<05:09, 1.15it/s] 78%\|███████▊ \| 1265/1621 [22:05<05:06, 1.16it/s] 78%\|███████▊ \| 1266/1621 [22:06<05:04, 1.17it/s] 78%\|███████▊ \| 1267/1621 [22:07<05:04, 1.16it/s] 78%\|███████▊ \| 1268/1621 [22:08<05:05, 1.16it/s] 78%\|███████▊ \| 1269/1621 [22:08<05:03, 1.16it/s] 78%\|███████▊ \| 1270/1621 [22:09<05:04, 1.15it/s] 78%\|██████�
	0: {'loss': 0.2891, 'grad_norm': 0.3171141586431303, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.79}
	0: �▊ \| 1270/1621 [22:09<05:04, 1.15it/s] 78%\|███████▊ \| 1271/1621 [22:10<05:26, 1.07it/s] 78%\|███████▊ \| 1272/1621 [22:11<05:16, 1.10it/s] 79%\|███████▊ \| 1273/1621 [22:12<05:09, 1.12it/s] 79%\|███████▊ \| 1274/1621 [22:13<05:04, 1.14it/s] 79%\|███████▊ \| 1275/1621 [22:14<05:07, 1.13it/s] 79%\|███████▊ \| 1276/1621 [22:15<05:05, 1.13it/s] 79%\|███████▉ \| 1277/1621 [22:16<05:02, 1.14it/s] 79%\|███████▉ \| 1278/1621 [22:16<04:58, 1.15it/s] 79%\|███████▉ \| 1279/1621 [22:17<04:56, 1.15it/s] 79%\|███████▉ \| 1280/1621 [22:18<04:54, 1.16it/s] 79%\|███████▉ \| 1280/1621 [22:18<04:54, 1.16it/s] 79%\|███████▉ \| 1281/1621 [22:19<04:53, 1.16it/s] 79%\|███████▉ \| 1282/1621 [22:20<04:57, 1.14it/s] 79%\|███████▉ \| 1283/1621 [
	0: {'loss': 0.2838, 'grad_norm': 0.2925938234274559, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.8}
	0: 22:21<05:00, 1.13it/s] 79%\|███████▉ \| 1284/1621 [22:22<04:55, 1.14it/s] 79%\|███████▉ \| 1285/1621 [22:23<05:04, 1.10it/s] 79%\|███████▉ \| 1286/1621 [22:24<05:00, 1.11it/s] 79%\|███████▉ \| 1287/1621 [22:24<05:01, 1.11it/s] 79%\|███████▉ \| 1288/1621 [22:25<04:57, 1.12it/s] 80%\|███████▉ \| 1289/1621 [22:26<04:53, 1.13it/s] 80%\|███████▉ \| 1290/1621 [22:27<04:50, 1.14it/s] 80%\|███████▉ \| 1290/1621 [22:27<04:50, 1.14it/s] 80%\|███████▉ \| 1291/1621 [22:28<04:59, 1.10it/s] 80%\|███████▉ \| 1292/1621 [22:29<04:52, 1.12it/s] 80%\|███████▉ \| 1293/1621 [22:30<04:48, 1.14it/s] 80%\|███████▉ \| 1294/1621 [22:31<04:45, 1.15it/s] 80%\|███████▉ \| 1295/1621 [22:31<04:43, 1.15it/s] 80%\|███████▉ \| 1296/1621 [22:32<04:46, 1.14i
	0: {'loss': 0.2905, 'grad_norm': 0.30092146988101065, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.8}
	0: [2025-09-02 18:58:46,711] [INFO] [axolotl.core.trainers.base._save:613] [PID:801836] [RANK:0] Saving model checkpoint to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0/checkpoint-1300[39m
	0: [2025-09-02 18:58:49,199] [INFO] [axolotl.core.trainers.base._save:662] [PID:801836] [RANK:0] Saving Trainer.data_collator.tokenizer by default as Trainer.processing_class is `None`[39m
	0: t/s] 80%\|████████ \| 1297/1621 [22:33<04:44, 1.14it/s] 80%\|████████ \| 1298/1621 [22:34<04:41, 1.15it/s] 80%\|████████ \| 1299/1621 [22:35<04:39, 1.15it/s] 80%\|████████ \| 1300/1621 [22:36<04:38, 1.15it/s] 80%\|████████ \| 1300/1621 [22:36<04:38, 1.15it/s] 80%\|████████ \| 1301/1621 [22:41<11:51, 2.22s/it] 80%\|████████ \| 1302/1621 [22:42<09:48, 1.85s/it] 80%\|████████ \| 1303/1621 [22:43<08:13, 1.55s/it] 80%\|████████ \| 1304/1621 [22:44<07:05, 1.34s/it] 81%\|████████ \| 1305/1621 [22:45<06:37, 1.26s/it] 81%\|████████ \| 1306/1621 [22:46<05:59, 1.14s/it] 81%\|████████ \| 1307/1621 [22:47<05:32, 1.06s/it] 81%\|████████ \| 1308/1621 [22:48<05:13, 1.00s/it] 81%\|████████ \| 1309/1621 [22:48<04:58, 1.05it/s] 81%\|███
	0: {'loss': 0.2876, 'grad_norm': 0.30256147814489603, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.81}
	0: {'loss': 0.2903, 'grad_norm': 0.29325662274638065, 'learning_rate': 7e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.81}
	0: █████ \| 1310/1621 [22:49<04:47, 1.08it/s] 81%\|████████ \| 1310/1621 [22:49<04:47, 1.08it/s] 81%\|████████ \| 1311/1621 [22:50<04:40, 1.10it/s] 81%\|████████ \| 1312/1621 [22:51<04:36, 1.12it/s] 81%\|████████ \| 1313/1621 [22:52<04:31, 1.14it/s] 81%\|████████ \| 1314/1621 [22:53<04:28, 1.14it/s] 81%\|████████ \| 1315/1621 [22:54<04:26, 1.15it/s] 81%\|████████ \| 1316/1621 [22:54<04:25, 1.15it/s] 81%\|████████ \| 1317/1621 [22:55<04:24, 1.15it/s] 81%\|████████▏ \| 1318/1621 [22:56<04:22, 1.15it/s] 81%\|████████▏ \| 1319/1621 [22:57<04:21, 1.15it/s] 81%\|████████▏ \| 1320/1621 [22:58<04:21, 1.15it/s] 81%\|████████▏ \| 1320/1621 [22:58<04:21, 1.15it/s] 81%\|████████▏ \| 1321/
	0: {'loss': 0.2853, 'grad_norm': 0.32226439329094486, 'learning_rate': 6.9889525066359386e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.82}
	0: 1621 [22:59<04:22, 1.14it/s] 82%\|████████▏ \| 1322/1621 [23:00<04:20, 1.15it/s] 82%\|████████▏ \| 1323/1621 [23:00<04:18, 1.15it/s] 82%\|████████▏ \| 1324/1621 [23:01<04:17, 1.15it/s] 82%\|████████▏ \| 1325/1621 [23:02<04:15, 1.16it/s] 82%\|████████▏ \| 1326/1621 [23:03<04:13, 1.16it/s] 82%\|████████▏ \| 1327/1621 [23:04<04:13, 1.16it/s] 82%\|████████▏ \| 1328/1621 [23:05<04:18, 1.13it/s] 82%\|████████▏ \| 1329/1621 [23:06<04:15, 1.14it/s] 82%\|████████▏ \| 1330/1621 [23:07<04:14, 1.14it/s] 82%\|████████▏ \| 1330/1621 [23:07<04:14, 1.14it/s] 82%\|████████▏ \| 1331/1621 [23:07<04:13, 1.15it/s] 82%\|████████▏ \| 1332/1621 [23:08<04:14, 1.14it/s] 82%\|████████▏ \| 1333/1621 [23:09<04:11, 1.15it/s] 82%\|████████�
	0: {'loss': 0.2837, 'grad_norm': 0.30545478926692277, 'learning_rate': 6.94420483979537e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.83}
	0: � \| 1334/1621 [23:10<04:25, 1.08it/s] 82%\|████████▏ \| 1335/1621 [23:11<04:23, 1.09it/s] 82%\|████████▏ \| 1336/1621 [23:12<04:16, 1.11it/s] 82%\|████████▏ \| 1337/1621 [23:13<04:11, 1.13it/s] 83%\|████████▎ \| 1338/1621 [23:14<04:08, 1.14it/s] 83%\|████████▎ \| 1339/1621 [23:15<04:12, 1.11it/s] 83%\|████████▎ \| 1340/1621 [23:16<04:08, 1.13it/s] 83%\|████████▎ \| 1340/1621 [23:16<04:08, 1.13it/s] 83%\|████████▎ \| 1341/1621 [23:16<04:05, 1.14it/s] 83%\|████████▎ \| 1342/1621 [23:17<04:05, 1.13it/s] 83%\|████████▎ \| 1343/1621 [23:18<04:12, 1.10it/s] 83%\|████████▎ \| 1344/1621 [23:19<04:06, 1.12it/s] 83%\|████████▎ \| 1345/1621 [23:20<04:03, 1.13it/s] 83%\|████████▎ \| 1346/1621 [23:21<04:00, 1.14it/s] 83%\|█████�
	0: {'loss': 0.2878, 'grad_norm': 0.3035523369868482, 'learning_rate': 6.865556417226012e-06, 'memory/max_mem_active(gib)': 32.71, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.83}
	0: �██▎ \| 1347/1621 [23:22<03:58, 1.15it/s] 83%\|████████▎ \| 1348/1621 [23:23<03:56, 1.16it/s] 83%\|████████▎ \| 1349/1621 [23:23<03:53, 1.17it/s] 83%\|████████▎ \| 1350/1621 [23:24<03:51, 1.17it/s] 83%\|████████▎ \| 1350/1621 [23:24<03:51, 1.17it/s] 83%\|████████▎ \| 1351/1621 [23:25<04:13, 1.06it/s] 83%\|████████▎ \| 1352/1621 [23:26<04:05, 1.10it/s] 83%\|████████▎ \| 1353/1621 [23:27<03:59, 1.12it/s] 84%\|████████▎ \| 1354/1621 [23:28<03:57, 1.13it/s] 84%\|████████▎ \| 1355/1621 [23:29<03:52, 1.14it/s] 84%\|████████▎ \| 1356/1621 [23:30<03:51, 1.15it/s] 84%\|████████▎ \| 1357/1621 [23:31<03:52, 1.13it/s] 84%\|████████▍ \| 1358/1621 [23:31<03:49, 1.15it/s] 84%\|████████▍ \| 1359/1621 [23:32<04:03, 1.08it/s] 84%\|██�
	0: {'loss': 0.285, 'grad_norm': 0.3004133597396245, 'learning_rate': 6.7538689275037765e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.84}
	0: {'loss': 0.2822, 'grad_norm': 0.28270553303909407, 'learning_rate': 6.61036604213817e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.85}
	0: �█████▍ \| 1360/1621 [23:33<03:57, 1.10it/s] 84%\|████████▍ \| 1360/1621 [23:33<03:57, 1.10it/s] 84%\|████████▍ \| 1361/1621 [23:34<03:52, 1.12it/s] 84%\|████████▍ \| 1362/1621 [23:35<03:47, 1.14it/s] 84%\|████████▍ \| 1363/1621 [23:36<03:46, 1.14it/s] 84%\|████████▍ \| 1364/1621 [23:37<03:54, 1.09it/s] 84%\|████████▍ \| 1365/1621 [23:38<03:49, 1.12it/s] 84%\|████████▍ \| 1366/1621 [23:39<03:48, 1.12it/s] 84%\|████████▍ \| 1367/1621 [23:39<03:44, 1.13it/s] 84%\|████████▍ \| 1368/1621 [23:40<03:45, 1.12it/s] 84%\|████████▍ \| 1369/1621 [23:41<03:41, 1.14it/s] 85%\|████████▍ \| 1370/1621 [23:42<03:38, 1.15it/s] 85%\|████████▍ \| 1370/1621 [23:42<03:38, 1.15it/s] 85%\|█████�
	0: {'loss': 0.2879, 'grad_norm': 0.3047147743418751, 'learning_rate': 6.436620008771168e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.85}
	0: ��██▍ \| 1371/1621 [23:43<03:47, 1.10it/s] 85%\|████████▍ \| 1372/1621 [23:44<03:43, 1.11it/s] 85%\|████████▍ \| 1373/1621 [23:45<03:40, 1.13it/s] 85%\|████████▍ \| 1374/1621 [23:46<03:37, 1.13it/s] 85%\|████████▍ \| 1375/1621 [23:47<03:35, 1.14it/s] 85%\|████████▍ \| 1376/1621 [23:48<03:42, 1.10it/s] 85%\|████████▍ \| 1377/1621 [23:48<03:42, 1.10it/s] 85%\|████████▌ \| 1378/1621 [23:49<03:38, 1.11it/s] 85%\|████████▌ \| 1379/1621 [23:50<03:34, 1.13it/s] 85%\|████████▌ \| 1380/1621 [23:51<03:31, 1.14it/s] 85%\|████████▌ \| 1380/1621 [23:51<03:31, 1.14it/s] 85%\|████████▌ \| 1381/1621 [23:52<03:29, 1.15it/s] 85%\|████████▌ \| 1382/1621 [23:53<03:27, 1.15it/s] 85%\|████████▌ \| 1383/1621 [23:54<03:29, 1.13it/s] 85%\|██�
	0: {'loss': 0.2813, 'grad_norm': 0.29257326468017114, 'learning_rate': 6.234534425303033e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.86}
	0: ��█████▌ \| 1384/1621 [23:55<03:27, 1.14it/s] 85%\|████████▌ \| 1385/1621 [23:55<03:26, 1.14it/s] 86%\|████████▌ \| 1386/1621 [23:56<03:27, 1.13it/s] 86%\|████████▌ \| 1387/1621 [23:57<03:24, 1.14it/s] 86%\|████████▌ \| 1388/1621 [23:58<03:23, 1.14it/s] 86%\|████████▌ \| 1389/1621 [23:59<03:21, 1.15it/s] 86%\|████████▌ \| 1390/1621 [24:00<03:21, 1.15it/s] 86%\|████████▌ \| 1390/1621 [24:00<03:21, 1.15it/s] 86%\|████████▌ \| 1391/1621 [24:01<03:22, 1.13it/s] 86%\|████████▌ \| 1392/1621 [24:02<03:20, 1.14it/s] 86%\|████████▌ \| 1393/1621 [24:02<03:18, 1.15it/s] 86%\|████████▌ \| 1394/1621 [24:03<03:17, 1.15it/s] 86%\|████████▌ \| 1395/1621 [24:04<03:23, 1.11it/s] 86%\|████████▌ \| 1396/1621 [24:05<03:19, 1.13it/s] 86
	0: {'loss': 0.2892, 'grad_norm': 0.29126232634431454, 'learning_rate': 6.006323383675369e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.86}
	0: %\|████████▌ \| 1397/1621 [24:06<03:15, 1.14it/s] 86%\|████████▌ \| 1398/1621 [24:07<03:14, 1.15it/s] 86%\|████████▋ \| 1399/1621 [24:08<03:12, 1.15it/s] 86%\|████████▋ \| 1400/1621 [24:09<03:10, 1.16it/s] 86%\|████████▋ \| 1400/1621 [24:09<03:10, 1.16it/s] 86%\|████████▋ \| 1401/1621 [24:09<03:08, 1.16it/s] 86%\|████████▋ \| 1402/1621 [24:10<03:07, 1.17it/s] 87%\|████████▋ \| 1403/1621 [24:11<03:06, 1.17it/s] 87%\|████████▋ \| 1404/1621 [24:12<03:05, 1.17it/s] 87%\|████████▋ \| 1405/1621 [24:13<03:08, 1.15it/s] 87%\|████████▋ \| 1406/1621 [24:14<03:08, 1.14it/s] 87%\|████████▋ \| 1407/1621 [24:15<03:07, 1.14it/s] 87%\|████████▋ \| 1408/1621 [24:16<03:19, 1.07it/s] 87%\|████████▋ \| 1409/1621 [24:17<03:13, 1.09
	0: {'loss': 0.2822, 'grad_norm': 0.27692361391158343, 'learning_rate': 5.754487211816481e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.87}
	0: {'loss': 0.2783, 'grad_norm': 0.2902902504363524, 'learning_rate': 5.48178507952536e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.88}
	0: it/s] 87%\|████████▋ \| 1410/1621 [24:17<03:10, 1.11it/s] 87%\|████████▋ \| 1410/1621 [24:17<03:10, 1.11it/s] 87%\|████████▋ \| 1411/1621 [24:18<03:05, 1.13it/s] 87%\|████████▋ \| 1412/1621 [24:19<03:03, 1.14it/s] 87%\|████████▋ \| 1413/1621 [24:20<03:03, 1.13it/s] 87%\|████████▋ \| 1414/1621 [24:21<03:01, 1.14it/s] 87%\|████████▋ \| 1415/1621 [24:22<02:59, 1.15it/s] 87%\|████████▋ \| 1416/1621 [24:23<02:58, 1.15it/s] 87%\|████████▋ \| 1417/1621 [24:24<02:59, 1.14it/s] 87%\|████████▋ \| 1418/1621 [24:24<03:04, 1.10it/s] 88%\|████████▊ \| 1419/1621 [24:25<03:00, 1.12it/s] 88%\|████████▊ \| 1420/1621 [24:26<02:57, 1.13it/s] 88%\|████████▊ \| 1420/1621 [24:26<02:57, 1.13it/s] 8
	0: {'loss': 0.2819, 'grad_norm': 0.301471028456177, 'learning_rate': 5.191204768429978e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.88}
	0: 8%\|████████▊ \| 1421/1621 [24:27<02:59, 1.12it/s] 88%\|████████▊ \| 1422/1621 [24:28<02:55, 1.13it/s] 88%\|████████▊ \| 1423/1621 [24:29<02:54, 1.14it/s] 88%\|████████▊ \| 1424/1621 [24:30<02:52, 1.14it/s] 88%\|████████▊ \| 1425/1621 [24:31<02:50, 1.15it/s] 88%\|████████▊ \| 1426/1621 [24:31<02:48, 1.16it/s] 88%\|████████▊ \| 1427/1621 [24:32<02:47, 1.16it/s] 88%\|████████▊ \| 1428/1621 [24:33<02:51, 1.12it/s] 88%\|████████▊ \| 1429/1621 [24:34<02:51, 1.12it/s] 88%\|████████▊ \| 1430/1621 [24:35<02:48, 1.13it/s] 88%\|████████▊ \| 1430/1621 [24:35<02:48, 1.13it/s] 88%\|████████▊ \| 1431/1621 [24:36<02:45, 1.15it/s] 88%\|████████▊ \| 1432/1621 [24:37<02:44, 1.15it/s] 88%\|████████▊ \| 1433/1621 [24:38<02:53, 1.0
	0: {'loss': 0.2796, 'grad_norm': 0.30104829275719075, 'learning_rate': 4.885929937226537e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.89}
	0: 8it/s] 88%\|████████▊ \| 1434/1621 [24:39<02:48, 1.11it/s] 89%\|████████▊ \| 1435/1621 [24:40<02:55, 1.06it/s] 89%\|████████▊ \| 1436/1621 [24:41<02:51, 1.08it/s] 89%\|████████▊ \| 1437/1621 [24:42<02:54, 1.05it/s] 89%\|████████▊ \| 1438/1621 [24:42<02:49, 1.08it/s] 89%\|████████▉ \| 1439/1621 [24:43<02:44, 1.11it/s] 89%\|████████▉ \| 1440/1621 [24:44<02:41, 1.12it/s] 89%\|████████▉ \| 1440/1621 [24:44<02:41, 1.12it/s] 89%\|████████▉ \| 1441/1621 [24:45<02:38, 1.13it/s] 89%\|████████▉ \| 1442/1621 [24:46<02:37, 1.14it/s] 89%\|████████▉ \| 1443/1621 [24:47<02:35, 1.15it/s] 89%\|████████▉ \| 1444/1621 [24:48<02:35, 1.14it/s] 89%\|████████▉ \| 1445/1621 [24:48<02:35, 1.13it/s] 89%\|████████▉ \| 1446/1621 [24:49<02
	0: {'loss': 0.289, 'grad_norm': 0.32313381078930165, 'learning_rate': 4.569305240848566e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.89}
	0: :33, 1.14it/s] 89%\|████████▉ \| 1447/1621 [24:50<02:30, 1.15it/s] 89%\|████████▉ \| 1448/1621 [24:51<02:29, 1.16it/s] 89%\|████████▉ \| 1449/1621 [24:52<02:30, 1.14it/s] 89%\|████████▉ \| 1450/1621 [24:53<02:33, 1.12it/s] 89%\|████████▉ \| 1450/1621 [24:53<02:33, 1.12it/s] 90%\|████████▉ \| 1451/1621 [24:54<02:38, 1.08it/s] 90%\|████████▉ \| 1452/1621 [24:55<02:33, 1.10it/s] 90%\|████████▉ \| 1453/1621 [24:56<02:30, 1.12it/s] 90%\|████████▉ \| 1454/1621 [24:56<02:27, 1.13it/s] 90%\|████████▉ \| 1455/1621 [24:57<02:27, 1.13it/s] 90%\|████████▉ \| 1456/1621 [24:58<02:25, 1.14it/s] 90%\|████████▉ \| 1457/1621 [24:59<02:23, 1.14it/s] 90%\|████████▉ \| 1458/1621 [25:00<02:22, 1.15it/s] 90%\|█████████ \| 1459/1621
	0: {'loss': 0.2819, 'grad_norm': 0.28369782040728675, 'learning_rate': 4.244799685727559e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.9}
	0: {'loss': 0.285, 'grad_norm': 0.273992840302611, 'learning_rate': 3.9159686226325745e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.91}
	0: [25:01<02:23, 1.13it/s] 90%\|█████████ \| 1460/1621 [25:02<02:22, 1.13it/s] 90%\|█████████ \| 1460/1621 [25:02<02:22, 1.13it/s] 90%\|█████████ \| 1461/1621 [25:03<02:19, 1.14it/s] 90%\|█████████ \| 1462/1621 [25:03<02:18, 1.15it/s] 90%\|█████████ \| 1463/1621 [25:04<02:16, 1.16it/s] 90%\|█████████ \| 1464/1621 [25:05<02:15, 1.16it/s] 90%\|█████████ \| 1465/1621 [25:06<02:14, 1.16it/s] 90%\|█████████ \| 1466/1621 [25:07<02:14, 1.15it/s] 90%\|█████████ \| 1467/1621 [25:08<02:14, 1.15it/s] 91%\|█████████ \| 1468/1621 [25:09<02:13, 1.15it/s] 91%\|█████████ \| 1469/1621 [25:10<02:13, 1.14it/s] 91%\|█████████ \| 1470/1621 [25:10<02:12, 1.14it/s] 91%\|█████████ \| 1470/1621 [25:10<0
	0: {'loss': 0.2798, 'grad_norm': 0.303660187115468, 'learning_rate': 3.586414793503207e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.91}
	0: 2:12, 1.14it/s] 91%\|█████████ \| 1471/1621 [25:11<02:14, 1.12it/s] 91%\|█████████ \| 1472/1621 [25:12<02:11, 1.13it/s] 91%\|█████████ \| 1473/1621 [25:13<02:09, 1.15it/s] 91%\|█████████ \| 1474/1621 [25:14<02:07, 1.15it/s] 91%\|█████████ \| 1475/1621 [25:15<02:06, 1.15it/s] 91%\|█████████ \| 1476/1621 [25:16<02:05, 1.16it/s] 91%\|█████████ \| 1477/1621 [25:17<02:04, 1.16it/s] 91%\|█████████ \| 1478/1621 [25:17<02:04, 1.15it/s] 91%\|█████████ \| 1479/1621 [25:18<02:02, 1.15it/s] 91%\|█████████▏\| 1480/1621 [25:19<02:01, 1.16it/s] 91%\|█████████▏\| 1480/1621 [25:19<02:01, 1.16it/s] 91%\|█████████▏\| 1481/1621 [25:20<02:00, 1.16it/s] 91%\|█████████▏\| 1482/1621 [25:21<02:09, 1.07it/s] 91%\|█████████▏\|
	0: {'loss': 0.2826, 'grad_norm': 0.2972873589703875, 'learning_rate': 3.2597488590549667e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.92}
	0: 1483/1621 [25:22<02:05, 1.10it/s] 92%\|█████████▏\| 1484/1621 [25:23<02:02, 1.12it/s] 92%\|█████████▏\| 1485/1621 [25:24<02:05, 1.09it/s] 92%\|█████████▏\| 1486/1621 [25:25<02:07, 1.06it/s] 92%\|█████████▏\| 1487/1621 [25:26<02:02, 1.09it/s] 92%\|█████████▏\| 1488/1621 [25:26<01:58, 1.12it/s] 92%\|█████████▏\| 1489/1621 [25:27<01:57, 1.13it/s] 92%\|█████████▏\| 1490/1621 [25:28<01:55, 1.14it/s] 92%\|█████████▏\| 1490/1621 [25:28<01:55, 1.14it/s] 92%\|█████████▏\| 1491/1621 [25:29<01:53, 1.15it/s] 92%\|█████████▏\| 1492/1621 [25:30<01:51, 1.15it/s] 92%\|█████████▏\| 1493/1621 [25:31<01:50, 1.15it/s] 92%\|█████████▏\| 1494/1621 [25:32<01:50, 1.15it/s] 92%\|█████████▏\| 1495/1621 [25:33<01:49, 1.15it/s]
	0: {'loss': 0.2789, 'grad_norm': 0.28996847808960213, 'learning_rate': 2.9395498396249143e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.93}
	0: 92%\|█████████▏\| 1496/1621 [25:33<01:47, 1.16it/s] 92%\|█████████▏\| 1497/1621 [25:34<01:49, 1.13it/s] 92%\|█████████▏\| 1498/1621 [25:35<01:48, 1.13it/s] 92%\|█████████▏\| 1499/1621 [25:36<01:46, 1.14it/s] 93%\|█████████▎\| 1500/1621 [25:37<01:44, 1.15it/s] 93%\|█████████▎\| 1500/1621 [25:37<01:44, 1.15it/s] 93%\|█████████▎\| 1501/1621 [25:38<01:45, 1.14it/s] 93%\|█████████▎\| 1502/1621 [25:39<01:44, 1.14it/s] 93%\|█████████▎\| 1503/1621 [25:40<01:43, 1.13it/s] 93%\|█████████▎\| 1504/1621 [25:40<01:42, 1.14it/s] 93%\|█████████▎\| 1505/1621 [25:41<01:40, 1.15it/s] 93%\|█████████▎\| 1506/1621 [25:42<01:39, 1.16it/s] 93%\|█████████▎\| 1507/1621 [25:43<01:39, 1.15it/s] 93%\|█████████�
	0: {'loss': 0.2848, 'grad_norm': 0.2966980326764614, 'learning_rate': 2.629325902675876e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.93}
	0: �\| 1508/1621 [25:44<01:37, 1.15it/s] 93%\|█████████▎\| 1509/1621 [25:45<01:36, 1.16it/s] 93%\|█████████▎\| 1510/1621 [25:46<01:35, 1.16it/s] 93%\|█████████▎\| 1510/1621 [25:46<01:35, 1.16it/s] 93%\|█████████▎\| 1511/1621 [25:46<01:34, 1.17it/s] 93%\|█████████▎\| 1512/1621 [25:47<01:33, 1.17it/s] 93%\|█████████▎\| 1513/1621 [25:48<01:38, 1.10it/s] 93%\|█████████▎\| 1514/1621 [25:49<01:43, 1.03it/s] 93%\|█████████▎\| 1515/1621 [25:50<01:39, 1.06it/s] 94%\|█████████▎\| 1516/1621 [25:51<01:36, 1.09it/s] 94%\|█████████▎\| 1517/1621 [25:52<01:34, 1.10it/s] 94%\|█████████▎\| 1518/1621 [25:53<01:33, 1.11it/s] 94%\|█████████▎\| 1519/1621 [25:54<01:30, 1.13it/s] 94%\|█████████▍\| 1520/1621 [25:55<01:28, 1.14it/
	0: {'loss': 0.2788, 'grad_norm': 0.2671505339196964, 'learning_rate': 2.3324759265795965e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.94}
	0: {'loss': 0.2899, 'grad_norm': 0.28163545935283396, 'learning_rate': 2.0522522617940406e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.94}
	0: s] 94%\|█████████▍\| 1520/1621 [25:55<01:28, 1.14it/s] 94%\|█████████▍\| 1521/1621 [25:56<01:27, 1.14it/s] 94%\|█████████▍\| 1522/1621 [25:56<01:25, 1.16it/s] 94%\|█████████▍\| 1523/1621 [25:57<01:24, 1.16it/s] 94%\|█████████▍\| 1524/1621 [25:58<01:25, 1.13it/s] 94%\|█████████▍\| 1525/1621 [25:59<01:23, 1.14it/s] 94%\|█████████▍\| 1526/1621 [26:00<01:23, 1.14it/s] 94%\|█████████▍\| 1527/1621 [26:01<01:22, 1.14it/s] 94%\|█████████▍\| 1528/1621 [26:02<01:22, 1.13it/s] 94%\|█████████▍\| 1529/1621 [26:03<01:24, 1.09it/s] 94%\|█████████▍\| 1530/1621 [26:04<01:22, 1.10it/s] 94%\|█████████▍\| 1530/1621 [26:04<01:22, 1.10it/s] 94%\|█████████▍\| 1531/1621 [26:
	0: {'loss': 0.2877, 'grad_norm': 0.29425238378168045, 'learning_rate': 1.7917250974311677e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.95}
	0: 04<01:20, 1.12it/s] 95%\|█████████▍\| 1532/1621 [26:05<01:18, 1.14it/s] 95%\|█████████▍\| 1533/1621 [26:06<01:16, 1.15it/s] 95%\|█████████▍\| 1534/1621 [26:07<01:15, 1.16it/s] 95%\|█████████▍\| 1535/1621 [26:08<01:18, 1.10it/s] 95%\|█████████▍\| 1536/1621 [26:09<01:15, 1.12it/s] 95%\|█████████▍\| 1537/1621 [26:10<01:15, 1.11it/s] 95%\|█████████▍\| 1538/1621 [26:11<01:14, 1.12it/s] 95%\|█████████▍\| 1539/1621 [26:11<01:12, 1.13it/s] 95%\|█████████▌\| 1540/1621 [26:12<01:10, 1.15it/s] 95%\|█████████▌\| 1540/1621 [26:12<01:10, 1.15it/s] 95%\|█████████▌\| 1541/1621 [26:13<01:12, 1.11it/s] 95%\|█████████▌\| 1542/1621 [26:14<01:10, 1.11it/s] 95%\|█████████▌\| 1543/1621 [26:15<01:09, 1.12it/s] 95%\|███
	0: {'loss': 0.2838, 'grad_norm': 0.2694113241241066, 'learning_rate': 1.5537488236225542e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.96}
	0: ██████▌\| 1544/1621 [26:16<01:07, 1.14it/s] 95%\|█████████▌\| 1545/1621 [26:17<01:06, 1.15it/s] 95%\|█████████▌\| 1546/1621 [26:18<01:05, 1.15it/s] 95%\|█████████▌\| 1547/1621 [26:19<01:05, 1.13it/s] 95%\|█████████▌\| 1548/1621 [26:19<01:04, 1.13it/s] 96%\|█████████▌\| 1549/1621 [26:20<01:02, 1.14it/s] 96%\|█████████▌\| 1550/1621 [26:21<01:01, 1.15it/s] 96%\|█████████▌\| 1550/1621 [26:21<01:01, 1.15it/s] 96%\|█████████▌\| 1551/1621 [26:22<01:00, 1.15it/s] 96%\|█████████▌\| 1552/1621 [26:23<01:01, 1.13it/s] 96%\|█████████▌\| 1553/1621 [26:24<00:59, 1.14it/s] 96%\|█████████▌\| 1554/1621 [26:25<00:59, 1.13it/s] 96%\|█████████▌\| 1555/1621 [26:26<00:57, 1.14it/s] 96%\|█████████▌\| 1556/1621 [2
	0: {'loss': 0.2834, 'grad_norm': 0.2739383897655746, 'learning_rate': 1.340930758223782e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.96}
	0: 6:26<00:56, 1.15it/s] 96%\|█████████▌\| 1557/1621 [26:27<00:55, 1.16it/s] 96%\|█████████▌\| 1558/1621 [26:28<00:55, 1.15it/s] 96%\|█████████▌\| 1559/1621 [26:29<00:55, 1.12it/s] 96%\|█████████▌\| 1560/1621 [26:30<00:53, 1.14it/s] 96%\|█████████▌\| 1560/1621 [26:30<00:53, 1.14it/s] 96%\|█████████▋\| 1561/1621 [26:31<00:54, 1.10it/s] 96%\|█████████▋\| 1562/1621 [26:32<00:53, 1.11it/s] 96%\|█████████▋\| 1563/1621 [26:33<00:51, 1.12it/s] 96%\|█████████▋\| 1564/1621 [26:33<00:50, 1.14it/s] 97%\|█████████▋\| 1565/1621 [26:34<00:48, 1.15it/s] 97%\|█████████▋\| 1566/1621 [26:35<00:47, 1.15it/s] 97%\|█████████▋\| 1567/1621 [26:36<00:46, 1.16it/s] 97%\|█████████▋\| 1568/1621 [26:37<00:48, 1.09it/s] 97%\|██�
	0: {'loss': 0.2845, 'grad_norm': 0.2793598127461013, 'learning_rate': 1.1556025804944043e-06, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.97}
	0: ��██████▋\| 1569/1621 [26:38<00:47, 1.10it/s] 97%\|█████████▋\| 1570/1621 [26:39<00:45, 1.13it/s] 97%\|█████████▋\| 1570/1621 [26:39<00:45, 1.13it/s] 97%\|█████████▋\| 1571/1621 [26:40<00:43, 1.14it/s] 97%\|█████████▋\| 1572/1621 [26:41<00:43, 1.13it/s] 97%\|█████████▋\| 1573/1621 [26:41<00:41, 1.15it/s] 97%\|█████████▋\| 1574/1621 [26:42<00:40, 1.16it/s] 97%\|█████████▋\| 1575/1621 [26:43<00:39, 1.16it/s] 97%\|█████████▋\| 1576/1621 [26:44<00:39, 1.15it/s] 97%\|█████████▋\| 1577/1621 [26:45<00:38, 1.15it/s] 97%\|█████████▋\| 1578/1621 [26:46<00:37, 1.15it/s] 97%\|█████████▋\| 1579/1621 [26:47<00:36, 1.16it/s] 97%\|█████████▋\| 1580/1621 [26:47<00:35, 1.16it/s]
	0: {'loss': 0.2878, 'grad_norm': 0.27156348472491937, 'learning_rate': 9.99794784732039e-07, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.97}
	0: {'loss': 0.2781, 'grad_norm': 0.2638303305592663, 'learning_rate': 8.752144337519349e-07, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.98}
	0: 97%\|█████████▋\| 1580/1621 [26:47<00:35, 1.16it/s] 98%\|█████████▊\| 1581/1621 [26:49<00:36, 1.08it/s] 98%\|█████████▊\| 1582/1621 [26:49<00:35, 1.11it/s] 98%\|█████████▊\| 1583/1621 [26:50<00:34, 1.10it/s] 98%\|█████████▊\| 1584/1621 [26:51<00:33, 1.12it/s] 98%\|█████████▊\| 1585/1621 [26:52<00:31, 1.14it/s] 98%\|█████████▊\| 1586/1621 [26:53<00:30, 1.15it/s] 98%\|█████████▊\| 1587/1621 [26:54<00:29, 1.15it/s] 98%\|█████████▊\| 1588/1621 [26:55<00:28, 1.16it/s] 98%\|█████████▊\| 1589/1621 [26:55<00:27, 1.16it/s] 98%\|█████████▊\| 1590/1621 [26:56<00:27, 1.14it/s] 98%\|█████████▊\| 1590/1621 [26:56<00:27, 1.14it/s] 98%\|█████████▊\| 1591/1621 [26:57<00:26, 1.14it/s] 98%\|████████
	0: {'loss': 0.2831, 'grad_norm': 0.27452303324042504, 'learning_rate': 7.832264559495451e-07, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.99}
	0: █▊\| 1592/1621 [26:58<00:25, 1.15it/s] 98%\|█████████▊\| 1593/1621 [26:59<00:24, 1.15it/s] 98%\|█████████▊\| 1594/1621 [27:00<00:23, 1.15it/s] 98%\|█████████▊\| 1595/1621 [27:01<00:22, 1.16it/s] 98%\|█████████▊\| 1596/1621 [27:02<00:22, 1.10it/s] 99%\|█████████▊\| 1597/1621 [27:03<00:21, 1.12it/s] 99%\|█████████▊\| 1598/1621 [27:03<00:20, 1.12it/s] 99%\|█████████▊\| 1599/1621 [27:04<00:19, 1.13it/s] 99%\|█████████▊\| 1600/1621 [27:05<00:18, 1.12it/s] 99%\|█████████▊\| 1600/1621 [27:05<00:18, 1.12it/s] 99%\|█████████▉\| 1601/1621 [27:06<00:17, 1.14it/s] 99%\|█████████▉\| 1602/1621 [27:07<00:16, 1.14it/s] 99%\|█████████▉\| 1603/1621 [27:08<00:15, 1.15it/s] 99%\|█████████▉\| 1604/1621 [27:09<00:14, 1.
	0: {'loss': 0.2921, 'grad_norm': 0.2764208630314015, 'learning_rate': 7.248386908593951e-07, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 0.99}
	0: 14it/s] 99%\|█████████▉\| 1605/1621 [27:10<00:13, 1.15it/s] 99%\|█████████▉\| 1606/1621 [27:10<00:13, 1.15it/s] 99%\|█████████▉\| 1607/1621 [27:11<00:12, 1.16it/s] 99%\|█████████▉\| 1608/1621 [27:12<00:11, 1.16it/s] 99%\|█████████▉\| 1609/1621 [27:13<00:10, 1.15it/s] 99%\|█████████▉\| 1610/1621 [27:14<00:09, 1.15it/s] 99%\|█████████▉\| 1610/1621 [27:14<00:09, 1.15it/s] 99%\|█████████▉\| 1611/1621 [27:15<00:08, 1.15it/s] 99%\|█████████▉\| 1612/1621 [27:16<00:07, 1.16it/s] 100%\|█████████▉\| 1613/1621 [27:16<00:06, 1.15it/s] 100%\|█████████▉\| 1614/1621 [27:17<00:06, 1.10it/s] 100%\|█████████▉\| 1615/1621 [27:18<00:05, 1.10it/s] 100%\|█████████▉\| 1616/1621 [27:19<00:04, 1.12it/s] 100%\|███████�
	0: {'loss': 0.283, 'grad_norm': 0.26571883451523387, 'learning_rate': 7.006908470542366e-07, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 1.0}
	0: [2025-09-02 19:03:37,520] [INFO] [axolotl.core.trainers.base._save:613] [PID:801836] [RANK:0] Saving model checkpoint to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0/checkpoint-1621[39m
	0: [2025-09-02 19:03:40,044] [INFO] [axolotl.core.trainers.base._save:662] [PID:801836] [RANK:0] Saving Trainer.data_collator.tokenizer by default as Trainer.processing_class is `None`[39m
	0: {'train_runtime': 1651.6754, 'train_samples_per_second': 15.703, 'train_steps_per_second': 0.981, 'train_loss': 0.2972346406744852, 'memory/max_mem_active(gib)': 32.97, 'memory/max_mem_allocated(gib)': 32.71, 'memory/device_mem_reserved(gib)': 38.82, 'epoch': 1.0}
	0: ��█▉\| 1617/1621 [27:20<00:03, 1.14it/s] 100%\|█████████▉\| 1618/1621 [27:21<00:02, 1.15it/s] 100%\|█████████▉\| 1619/1621 [27:22<00:01, 1.15it/s] 100%\|█████████▉\| 1620/1621 [27:23<00:00, 1.16it/s] 100%\|█████████▉\| 1620/1621 [27:23<00:00, 1.16it/s] 100%\|██████████\| 1621/1621 [27:27<00:00, 1.79s/it] 100%\|██████████\| 1621/1621 [27:31<00:00, 1.79s/it] 100%\|██████████\| 1621/1621 [27:31<00:00, 1.02s/it]
	0: [2025-09-02 19:03:41,703] [INFO] [axolotl.train.save_trained_model:228] [PID:801836] [RANK:0] Training completed! Saving trained model to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0.[39m
	0: [2025-09-02 19:03:42,150] [INFO] [axolotl.core.trainers.base._save:613] [PID:801836] [RANK:0] Saving model checkpoint to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0[39m
	0: [2025-09-02 19:03:44,635] [INFO] [axolotl.core.trainers.base._save:662] [PID:801836] [RANK:0] Saving Trainer.data_collator.tokenizer by default as Trainer.processing_class is `None`[39m
	0: [2025-09-02 19:03:44,822] [INFO] [axolotl.train.save_trained_model:350] [PID:801836] [RANK:0] Model successfully saved to /lustre/fswork/projects/rech/dgo/udv55np/math/Qwen3-235B-A22B/Qwen2.5-1.5B_ift/0[39m