hello-world-yoo commited on
Commit
336a9cb
·
verified ·
1 Parent(s): fc2e174

Upload folder using huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: nvidia/GR00T-N1.5-3B
4
+ tags:
5
+ - robotics
6
+ - gr00t
7
+ - vla
8
+ - manipulation
9
+ - diffusion-policy
10
+ datasets:
11
+ - hi-space/SO-ARM101-PICK-BANANA
12
+ library_name: gr00t
13
+ pipeline_tag: robotics
14
+ ---
15
+
16
+ # GR00T-N1.6-3B-Pick-Banana-Real
17
+
18
+ A fine-tuned version of [nvidia/GR00T-N1.5-3B](https://huggingface.co/nvidia/GR00T-N1.5-3B) for a banana pick-and-place task on a real SO-ARM101 robot, trained on the [hi-space/SO-ARM101-PICK-BANANA](https://huggingface.co/datasets/hi-space/SO-ARM101-PICK-BANANA) dataset.
19
+
20
+ ## Model Description
21
+
22
+ GR00T-N1.6 (Gr00tN1d6) is a vision-language-action (VLA) model for robot manipulation. This checkpoint is fine-tuned for a pick-and-place task where the robot picks up a banana and places it on a plate using a real SO-ARM101 robot arm.
23
+
24
+ - **Architecture:** Gr00tN1d6 with Eagle-Block2A-2B-v2 vision-language backbone + diffusion policy action head
25
+ - **Base model:** nvidia/GR00T-N1.5-3B
26
+ - **Task:** Pick banana and place on plate (real robot)
27
+ - **Robot:** SO-ARM101
28
+ - **Action horizon:** 50 steps
29
+ - **Inference timesteps:** 4 (diffusion)
30
+ - **Model dtype:** bfloat16
31
+
32
+ ### Fine-tuning Configuration
33
+
34
+ | Parameter | Value |
35
+ |-----------|-------|
36
+ | Tuned components | Diffusion model, projector, top 4 LLM layers, VL-LN |
37
+ | Frozen components | Vision encoder, LLM backbone |
38
+ | Training steps | 6000 |
39
+ | Epochs | 1 |
40
+ | Final training loss | ~0.017 |
41
+ | Action representation | Relative actions |
42
+ | Attention | Flash Attention 2 |
43
+
44
+ ## Training Details
45
+
46
+ - **Dataset:** [hi-space/SO-ARM101-PICK-BANANA](https://huggingface.co/datasets/hi-space/SO-ARM101-PICK-BANANA)
47
+ - **Episodes:** 58
48
+ - **Total frames:** 35,859
49
+ - **Max steps:** 6000 (1 epoch)
50
+ - **Loss curve:** Started at ~1.13, converged to ~0.017
51
+
52
+ ## Usage
53
+
54
+ ```python
55
+ from gr00t.model.gr00t_n1 import GR00TPolicy
56
+
57
+ policy = GR00TPolicy.from_pretrained("hi-space/GR00T-N1.6-3B-Pick-Banana-Real")
58
+ ```
59
+
60
+ Refer to the [NVIDIA Isaac GR00T](https://github.com/NVIDIA/Isaac-GR00T) repository for full inference and deployment instructions.
61
+
62
+ ## Intended Use
63
+
64
+ This model is fine-tuned for a real-world robotic banana pick-and-place task using the SO-ARM101 robot arm. It is trained on real robot demonstrations and intended for deployment on the same hardware setup.
65
+
66
+ ## License
67
+
68
+ This model inherits the license from the base model [nvidia/GR00T-N1.5-3B](https://huggingface.co/nvidia/GR00T-N1.5-3B). Please refer to NVIDIA's terms for usage restrictions.
config.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "action_horizon": 50,
3
+ "add_pos_embed": true,
4
+ "apply_sincos_state_encoding": true,
5
+ "architectures": [
6
+ "Gr00tN1d6"
7
+ ],
8
+ "attn_dropout": 0.2,
9
+ "attn_implementation": null,
10
+ "backbone_embedding_dim": 2048,
11
+ "backbone_model_type": "eagle",
12
+ "backbone_trainable_params_fp32": true,
13
+ "collator_overwrite_image_inputs": false,
14
+ "color_jitter_params": {
15
+ "brightness": 0.1,
16
+ "contrast": 0.1,
17
+ "hue": 0.1,
18
+ "saturation": 0.1
19
+ },
20
+ "crop_fraction": 0.95,
21
+ "diffusion_model_cfg": {
22
+ "attention_head_dim": 48,
23
+ "dropout": 0.2,
24
+ "final_dropout": true,
25
+ "interleave_self_attention": true,
26
+ "norm_type": "ada_norm",
27
+ "num_attention_heads": 32,
28
+ "num_layers": 32,
29
+ "output_dim": 1024,
30
+ "positional_embeddings": null
31
+ },
32
+ "eagle_collator": true,
33
+ "formalize_language": true,
34
+ "gemma_collator": false,
35
+ "hidden_size": 1024,
36
+ "image_crop_size": null,
37
+ "image_target_size": null,
38
+ "input_embedding_dim": 1536,
39
+ "load_bf16": true,
40
+ "max_action_dim": 128,
41
+ "max_num_embodiments": 32,
42
+ "max_seq_len": 1024,
43
+ "max_state_dim": 128,
44
+ "model_dtype": "bfloat16",
45
+ "model_name": "nvidia/Eagle-Block2A-2B-v2",
46
+ "model_type": "Gr00tN1d6",
47
+ "noise_beta_alpha": 1.5,
48
+ "noise_beta_beta": 1.0,
49
+ "noise_s": 0.999,
50
+ "num_inference_timesteps": 4,
51
+ "num_timestep_buckets": 1000,
52
+ "random_rotation_angle": null,
53
+ "reproject_vision": false,
54
+ "select_layer": 16,
55
+ "shortest_image_edge": 256,
56
+ "state_dropout_prob": 0.0,
57
+ "torch_dtype": "bfloat16",
58
+ "transformers_version": "4.51.3",
59
+ "tune_diffusion_model": true,
60
+ "tune_llm": false,
61
+ "tune_projector": true,
62
+ "tune_top_llm_layers": 4,
63
+ "tune_visual": false,
64
+ "tune_vlln": true,
65
+ "use_albumentations_transforms": true,
66
+ "use_alternate_vl_dit": true,
67
+ "use_flash_attention": true,
68
+ "use_relative_action": true,
69
+ "use_vlln": true
70
+ }
embodiment_id.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "robocasa_panda_omron": 13,
3
+ "gr1": 20,
4
+ "behavior_r1_pro": 24,
5
+ "unitree_g1": 8,
6
+ "oxe_google": 0,
7
+ "oxe_widowx": 1,
8
+ "libero_panda": 2,
9
+ "oxe_droid": 16,
10
+ "new_embodiment": 10
11
+ }
experiment_cfg/conf.yaml ADDED
@@ -0,0 +1,206 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ load_config_path: null
2
+ model:
3
+ model_type: Gr00tN1d6
4
+ model_dtype: bfloat16
5
+ model_name: nvidia/Eagle-Block2A-2B-v2
6
+ backbone_model_type: eagle
7
+ model_revision: null
8
+ tune_top_llm_layers: 4
9
+ backbone_embedding_dim: 2048
10
+ tune_llm: false
11
+ tune_visual: false
12
+ select_layer: 16
13
+ reproject_vision: false
14
+ use_flash_attention: true
15
+ load_bf16: false
16
+ collator_overwrite_image_inputs: false
17
+ eagle_collator: false
18
+ backbone_trainable_params_fp32: true
19
+ image_crop_size: null
20
+ image_target_size: null
21
+ shortest_image_edge: 256
22
+ crop_fraction: 0.95
23
+ random_rotation_angle: null
24
+ color_jitter_params: null
25
+ use_albumentations_transforms: true
26
+ extra_augmentation_config: null
27
+ formalize_language: true
28
+ apply_sincos_state_encoding: false
29
+ use_relative_action: true
30
+ max_state_dim: 29
31
+ max_action_dim: 29
32
+ action_horizon: 16
33
+ hidden_size: 1024
34
+ input_embedding_dim: 1536
35
+ add_pos_embed: true
36
+ attn_dropout: 0.2
37
+ use_vlln: true
38
+ max_seq_len: 1024
39
+ use_alternate_vl_dit: true
40
+ attend_text_every_n_blocks: 2
41
+ diffusion_model_cfg:
42
+ positional_embeddings: null
43
+ num_layers: 32
44
+ num_attention_heads: 32
45
+ attention_head_dim: 48
46
+ norm_type: ada_norm
47
+ dropout: 0.2
48
+ final_dropout: true
49
+ output_dim: 1024
50
+ interleave_self_attention: true
51
+ num_inference_timesteps: 4
52
+ noise_beta_alpha: 1.5
53
+ noise_beta_beta: 1.0
54
+ noise_s: 0.999
55
+ num_timestep_buckets: 1000
56
+ tune_projector: true
57
+ tune_diffusion_model: true
58
+ tune_vlln: true
59
+ state_dropout_prob: 0.0
60
+ state_additive_noise_scale: 0.0
61
+ max_num_embodiments: 32
62
+ data:
63
+ datasets:
64
+ - dataset_paths:
65
+ - /workspace/train
66
+ embodiment_tag: new_embodiment
67
+ mix_ratio: 1.0
68
+ dataset_type: physical_embodiment
69
+ val_dataset_path: null
70
+ modality_configs:
71
+ new_embodiment:
72
+ video:
73
+ delta_indices:
74
+ - 0
75
+ modality_keys:
76
+ - front
77
+ - wrist
78
+ sin_cos_embedding_keys: null
79
+ mean_std_embedding_keys: null
80
+ action_configs: null
81
+ state:
82
+ delta_indices:
83
+ - 0
84
+ modality_keys:
85
+ - single_arm
86
+ - gripper
87
+ sin_cos_embedding_keys: null
88
+ mean_std_embedding_keys: null
89
+ action_configs: null
90
+ action:
91
+ delta_indices:
92
+ - 0
93
+ - 1
94
+ - 2
95
+ - 3
96
+ - 4
97
+ - 5
98
+ - 6
99
+ - 7
100
+ - 8
101
+ - 9
102
+ - 10
103
+ - 11
104
+ - 12
105
+ - 13
106
+ - 14
107
+ - 15
108
+ modality_keys:
109
+ - single_arm
110
+ - gripper
111
+ sin_cos_embedding_keys: null
112
+ mean_std_embedding_keys: null
113
+ action_configs:
114
+ - rep: RELATIVE
115
+ type: NON_EEF
116
+ format: DEFAULT
117
+ state_key: null
118
+ - rep: ABSOLUTE
119
+ type: NON_EEF
120
+ format: DEFAULT
121
+ state_key: null
122
+ language:
123
+ delta_indices:
124
+ - 0
125
+ modality_keys:
126
+ - annotation.human.task_description
127
+ sin_cos_embedding_keys: null
128
+ mean_std_embedding_keys: null
129
+ action_configs: null
130
+ download_cache: false
131
+ shard_size: 1024
132
+ episode_sampling_rate: 0.1
133
+ num_shards_per_epoch: 100000
134
+ override_pretraining_statistics: false
135
+ mode: single_turn
136
+ random_chop: 0.0
137
+ mock_dataset_mode: false
138
+ shuffle: true
139
+ seed: 42
140
+ multiprocessing_context: fork
141
+ allow_padding: false
142
+ subsample_ratio: 1.0
143
+ image_crop_size:
144
+ - 244
145
+ - 244
146
+ image_target_size:
147
+ - 224
148
+ - 224
149
+ video_backend: torchcodec
150
+ training:
151
+ output_dir: /mnt/efs/gr00t/checkpoints/6f4f0bf8-fd17-47a5-a3c5-6bccc5bd08f8
152
+ experiment_name: null
153
+ max_steps: 6000
154
+ global_batch_size: 32
155
+ batch_size: null
156
+ gradient_accumulation_steps: 1
157
+ learning_rate: 0.0001
158
+ lr_scheduler_type: cosine
159
+ weight_decay: 1.0e-05
160
+ warmup_ratio: 0.05
161
+ warmup_steps: 0
162
+ max_grad_norm: 1.0
163
+ optim: adamw_torch
164
+ start_from_checkpoint: nvidia/GR00T-N1.6-3B
165
+ tf32: true
166
+ fp16: false
167
+ bf16: true
168
+ eval_bf16: true
169
+ logging_steps: 10
170
+ save_steps: 2000
171
+ save_total_limit: 5
172
+ save_vl_model: false
173
+ upload_checkpoints: false
174
+ upload_every: 1000
175
+ upload_last_n_checkpoints: 5
176
+ max_concurrent_uploads: 2
177
+ eval_strategy: 'no'
178
+ eval_steps: 500
179
+ eval_set_split_ratio: 0.1
180
+ eval_batch_size: 2
181
+ save_best_eval_metric_name: ''
182
+ save_best_eval_metric_greater_is_better: true
183
+ deepspeed_stage: 2
184
+ gradient_checkpointing: false
185
+ transformers_trust_remote_code: true
186
+ transformers_local_files_only: false
187
+ transformers_cache_dir: null
188
+ transformers_access_token: null
189
+ use_ddp: false
190
+ ddp_bucket_cap_mb: 100
191
+ num_gpus: 1
192
+ dataloader_num_workers: 4
193
+ remove_unused_columns: false
194
+ use_wandb: false
195
+ wandb_project: finetune-gr00t-n1d7
196
+ enable_profiling: false
197
+ max_retries: 3
198
+ assert_loss_less_than: null
199
+ add_rl_callback: false
200
+ enable_open_loop_eval: false
201
+ open_loop_eval_traj_ids:
202
+ - 0
203
+ open_loop_eval_steps_per_traj: 100
204
+ open_loop_eval_plot_indices: null
205
+ max_steps: 6000
206
+ save_steps: 2000
experiment_cfg/config.yaml ADDED
@@ -0,0 +1,236 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ !!python/object:gr00t.configs.base_config.Config
2
+ data: !!python/object:gr00t.configs.data.data_config.DataConfig
3
+ allow_padding: false
4
+ datasets:
5
+ - !!python/object:gr00t.configs.data.data_config.SingleDatasetConfig
6
+ dataset_paths:
7
+ - /workspace/train
8
+ dataset_type: physical_embodiment
9
+ embodiment_tag: new_embodiment
10
+ mix_ratio: 1.0
11
+ val_dataset_path: null
12
+ download_cache: false
13
+ episode_sampling_rate: 0.1
14
+ image_crop_size:
15
+ - 244
16
+ - 244
17
+ image_target_size:
18
+ - 224
19
+ - 224
20
+ mock_dataset_mode: false
21
+ modality_configs:
22
+ new_embodiment:
23
+ action: !!python/object:gr00t.data.types.ModalityConfig
24
+ action_configs:
25
+ - !!python/object:gr00t.data.types.ActionConfig
26
+ format: &id001 !!python/object/apply:gr00t.data.types.ActionFormat
27
+ - default
28
+ rep: !!python/object/apply:gr00t.data.types.ActionRepresentation
29
+ - relative
30
+ state_key: null
31
+ type: &id002 !!python/object/apply:gr00t.data.types.ActionType
32
+ - non_eef
33
+ - !!python/object:gr00t.data.types.ActionConfig
34
+ format: *id001
35
+ rep: !!python/object/apply:gr00t.data.types.ActionRepresentation
36
+ - absolute
37
+ state_key: null
38
+ type: *id002
39
+ delta_indices:
40
+ - 0
41
+ - 1
42
+ - 2
43
+ - 3
44
+ - 4
45
+ - 5
46
+ - 6
47
+ - 7
48
+ - 8
49
+ - 9
50
+ - 10
51
+ - 11
52
+ - 12
53
+ - 13
54
+ - 14
55
+ - 15
56
+ mean_std_embedding_keys: null
57
+ modality_keys:
58
+ - single_arm
59
+ - gripper
60
+ sin_cos_embedding_keys: null
61
+ language: !!python/object:gr00t.data.types.ModalityConfig
62
+ action_configs: null
63
+ delta_indices:
64
+ - 0
65
+ mean_std_embedding_keys: null
66
+ modality_keys:
67
+ - annotation.human.task_description
68
+ sin_cos_embedding_keys: null
69
+ state: !!python/object:gr00t.data.types.ModalityConfig
70
+ action_configs: null
71
+ delta_indices:
72
+ - 0
73
+ mean_std_embedding_keys: null
74
+ modality_keys:
75
+ - single_arm
76
+ - gripper
77
+ sin_cos_embedding_keys: null
78
+ video: !!python/object:gr00t.data.types.ModalityConfig
79
+ action_configs: null
80
+ delta_indices:
81
+ - 0
82
+ mean_std_embedding_keys: null
83
+ modality_keys:
84
+ - front
85
+ - wrist
86
+ sin_cos_embedding_keys: null
87
+ mode: single_turn
88
+ multiprocessing_context: fork
89
+ num_shards_per_epoch: 100000
90
+ override_pretraining_statistics: false
91
+ random_chop: 0.0
92
+ seed: 42
93
+ shard_size: 1024
94
+ shuffle: true
95
+ subsample_ratio: 1.0
96
+ video_backend: torchcodec
97
+ load_config_path: null
98
+ model: !!python/object:gr00t.configs.model.gr00t_n1d6.Gr00tN1d6Config
99
+ _attn_implementation_autoset: false
100
+ _attn_implementation_internal: null
101
+ _commit_hash: null
102
+ _name_or_path: ''
103
+ add_cross_attention: false
104
+ architectures: null
105
+ backbone_trainable_params_fp32: true
106
+ bad_words_ids: null
107
+ begin_suppress_tokens: null
108
+ bos_token_id: null
109
+ chunk_size_feed_forward: 0
110
+ color_jitter_params: null
111
+ cross_attention_hidden_size: null
112
+ decoder_start_token_id: null
113
+ diffusion_model_cfg:
114
+ attention_head_dim: 48
115
+ dropout: 0.2
116
+ final_dropout: true
117
+ interleave_self_attention: true
118
+ norm_type: ada_norm
119
+ num_attention_heads: 32
120
+ num_layers: 32
121
+ output_dim: 1024
122
+ positional_embeddings: null
123
+ diversity_penalty: 0.0
124
+ do_sample: false
125
+ early_stopping: false
126
+ encoder_no_repeat_ngram_size: 0
127
+ eos_token_id: null
128
+ exponential_decay_length_penalty: null
129
+ finetuning_task: null
130
+ forced_bos_token_id: null
131
+ forced_eos_token_id: null
132
+ id2label:
133
+ 0: LABEL_0
134
+ 1: LABEL_1
135
+ is_decoder: false
136
+ is_encoder_decoder: false
137
+ label2id:
138
+ LABEL_0: 0
139
+ LABEL_1: 1
140
+ length_penalty: 1.0
141
+ load_bf16: false
142
+ max_length: 20
143
+ min_length: 0
144
+ no_repeat_ngram_size: 0
145
+ num_beam_groups: 1
146
+ num_beams: 1
147
+ num_return_sequences: 1
148
+ output_attentions: false
149
+ output_hidden_states: false
150
+ output_scores: false
151
+ pad_token_id: null
152
+ prefix: null
153
+ problem_type: null
154
+ pruned_heads: {}
155
+ random_rotation_angle: null
156
+ remove_invalid_values: false
157
+ repetition_penalty: 1.0
158
+ reproject_vision: false
159
+ return_dict: true
160
+ return_dict_in_generate: false
161
+ sep_token_id: null
162
+ state_dropout_prob: 0.0
163
+ suppress_tokens: null
164
+ task_specific_params: null
165
+ temperature: 1.0
166
+ tf_legacy_loss: false
167
+ tie_encoder_decoder: false
168
+ tie_word_embeddings: true
169
+ tokenizer_class: null
170
+ top_k: 50
171
+ top_p: 1.0
172
+ torch_dtype: null
173
+ torchscript: false
174
+ transformers_version: null
175
+ tune_diffusion_model: true
176
+ tune_llm: false
177
+ tune_projector: true
178
+ tune_visual: false
179
+ typical_p: 1.0
180
+ use_bfloat16: false
181
+ use_relative_action: true
182
+ training: !!python/object:gr00t.configs.training.training_config.TrainingConfig
183
+ add_rl_callback: false
184
+ assert_loss_less_than: null
185
+ batch_size: null
186
+ bf16: true
187
+ dataloader_num_workers: 4
188
+ ddp_bucket_cap_mb: 100
189
+ deepspeed_stage: 2
190
+ enable_open_loop_eval: false
191
+ enable_profiling: false
192
+ eval_batch_size: 2
193
+ eval_bf16: true
194
+ eval_set_split_ratio: 0.1
195
+ eval_steps: 500
196
+ eval_strategy: 'no'
197
+ experiment_name: null
198
+ fp16: false
199
+ global_batch_size: 32
200
+ gradient_accumulation_steps: 1
201
+ gradient_checkpointing: false
202
+ learning_rate: 0.0001
203
+ logging_steps: 10
204
+ lr_scheduler_type: cosine
205
+ max_concurrent_uploads: 2
206
+ max_grad_norm: 1.0
207
+ max_retries: 3
208
+ max_steps: 6000
209
+ num_gpus: 1
210
+ open_loop_eval_plot_indices: null
211
+ open_loop_eval_steps_per_traj: 100
212
+ open_loop_eval_traj_ids:
213
+ - 0
214
+ optim: adamw_torch
215
+ output_dir: /mnt/efs/gr00t/checkpoints/6f4f0bf8-fd17-47a5-a3c5-6bccc5bd08f8
216
+ remove_unused_columns: false
217
+ save_best_eval_metric_greater_is_better: true
218
+ save_best_eval_metric_name: ''
219
+ save_steps: 2000
220
+ save_total_limit: 5
221
+ save_vl_model: false
222
+ start_from_checkpoint: nvidia/GR00T-N1.6-3B
223
+ tf32: true
224
+ transformers_access_token: null
225
+ transformers_cache_dir: null
226
+ transformers_local_files_only: false
227
+ transformers_trust_remote_code: true
228
+ upload_checkpoints: false
229
+ upload_every: 1000
230
+ upload_last_n_checkpoints: 5
231
+ use_ddp: false
232
+ use_wandb: false
233
+ wandb_project: finetune-gr00t-n1d7
234
+ warmup_ratio: 0.05
235
+ warmup_steps: 0
236
+ weight_decay: 1.0e-05
experiment_cfg/dataset_statistics.json ADDED
@@ -0,0 +1,824 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "new_embodiment": {
3
+ "state": {
4
+ "single_arm": {
5
+ "min": [
6
+ -75.34066009521484,
7
+ -95.73626708984375,
8
+ -90.9011001586914,
9
+ -97.23076629638672,
10
+ -169.27471923828125
11
+ ],
12
+ "max": [
13
+ 61.18681335449219,
14
+ 63.56044006347656,
15
+ 93.27472686767578,
16
+ 96.70330047607422,
17
+ -48.131866455078125
18
+ ],
19
+ "mean": [
20
+ -15.921283721923828,
21
+ -36.2519416809082,
22
+ 43.81039047241211,
23
+ -4.216508865356445,
24
+ -151.60240173339844
25
+ ],
26
+ "std": [
27
+ 22.13420295715332,
28
+ 37.534183502197266,
29
+ 33.58217239379883,
30
+ 59.000938415527344,
31
+ 25.700382232666016
32
+ ],
33
+ "q01": [
34
+ -65.4065933227539,
35
+ -95.56044006347656,
36
+ -34.10988998413086,
37
+ -96.96703338623047,
38
+ -169.27471923828125
39
+ ],
40
+ "q99": [
41
+ 27.25274658203125,
42
+ 33.758243560791016,
43
+ 92.74725341796875,
44
+ 96.70330047607422,
45
+ -70.10989379882812
46
+ ]
47
+ },
48
+ "gripper": {
49
+ "min": [
50
+ 0.27913469076156616
51
+ ],
52
+ "max": [
53
+ 86.53175354003906
54
+ ],
55
+ "mean": [
56
+ 15.353877067565918
57
+ ],
58
+ "std": [
59
+ 14.16923999786377
60
+ ],
61
+ "q01": [
62
+ 0.27913469076156616
63
+ ],
64
+ "q99": [
65
+ 56.17585372924805
66
+ ]
67
+ }
68
+ },
69
+ "action": {
70
+ "single_arm": {
71
+ "min": [
72
+ -81.97801971435547,
73
+ -100.70330047607422,
74
+ -96.48351287841797,
75
+ -100.3956069946289,
76
+ -171.2967071533203
77
+ ],
78
+ "max": [
79
+ 61.758243560791016,
80
+ 62.989009857177734,
81
+ 97.97801971435547,
82
+ 111.82417297363281,
83
+ -47.42856979370117
84
+ ],
85
+ "mean": [
86
+ -15.838254928588867,
87
+ -37.501747131347656,
88
+ 43.07667541503906,
89
+ -4.388125419616699,
90
+ -152.02474975585938
91
+ ],
92
+ "std": [
93
+ 22.322818756103516,
94
+ 37.600399017333984,
95
+ 34.9411735534668,
96
+ 59.453250885009766,
97
+ 26.090877532958984
98
+ ],
99
+ "q01": [
100
+ -65.97801971435547,
101
+ -99.82417297363281,
102
+ -35.20878982543945,
103
+ -98.37362670898438,
104
+ -170.6813201904297
105
+ ],
106
+ "q99": [
107
+ 28.0,
108
+ 32.78417549133285,
109
+ 96.74725341796875,
110
+ 99.78022003173828,
111
+ -69.89714309692398
112
+ ]
113
+ },
114
+ "gripper": {
115
+ "min": [
116
+ 0.0
117
+ ],
118
+ "max": [
119
+ 86.90878295898438
120
+ ],
121
+ "mean": [
122
+ 11.563150405883789
123
+ ],
124
+ "std": [
125
+ 15.86936092376709
126
+ ],
127
+ "q01": [
128
+ 0.0
129
+ ],
130
+ "q99": [
131
+ 56.58783721923828
132
+ ]
133
+ }
134
+ },
135
+ "relative_action": {
136
+ "single_arm": {
137
+ "min": [
138
+ [
139
+ -17.098901748657227,
140
+ -22.901098251342773,
141
+ -31.252750396728516,
142
+ -30.32967185974121,
143
+ -33.31867980957031
144
+ ],
145
+ [
146
+ -21.05494499206543,
147
+ -27.736263275146484,
148
+ -37.494503021240234,
149
+ -36.92307662963867,
150
+ -40.175819396972656
151
+ ],
152
+ [
153
+ -25.010988235473633,
154
+ -32.483516693115234,
155
+ -43.20878982543945,
156
+ -43.164833068847656,
157
+ -46.857139587402344
158
+ ],
159
+ [
160
+ -28.703296661376953,
161
+ -37.14285659790039,
162
+ -48.13187026977539,
163
+ -49.14285659790039,
164
+ -53.362632751464844
165
+ ],
166
+ [
167
+ -32.0439567565918,
168
+ -41.71428680419922,
169
+ -52.96703338623047,
170
+ -55.120880126953125,
171
+ -59.164833068847656
172
+ ],
173
+ [
174
+ -35.29670333862305,
175
+ -46.19780349731445,
176
+ -57.27472686767578,
177
+ -60.92307662963867,
178
+ -64.70329284667969
179
+ ],
180
+ [
181
+ -38.28571319580078,
182
+ -50.32966995239258,
183
+ -60.879119873046875,
184
+ -66.37362670898438,
185
+ -69.53845977783203
186
+ ],
187
+ [
188
+ -41.27472686767578,
189
+ -54.46154022216797,
190
+ -64.04395294189453,
191
+ -71.73625946044922,
192
+ -74.02197265625
193
+ ],
194
+ [
195
+ -43.736263275146484,
196
+ -58.5054931640625,
197
+ -66.4175796508789,
198
+ -77.0988998413086,
199
+ -78.15383911132812
200
+ ],
201
+ [
202
+ -46.10988998413086,
203
+ -62.46154022216797,
204
+ -68.0879135131836,
205
+ -82.37362670898438,
206
+ -82.02197265625
207
+ ],
208
+ [
209
+ -48.13187026977539,
210
+ -66.5054931640625,
211
+ -69.05494689941406,
212
+ -87.64834594726562,
213
+ -85.89010620117188
214
+ ],
215
+ [
216
+ -50.153846740722656,
217
+ -70.28571319580078,
218
+ -69.23076629638672,
219
+ -92.57142639160156,
220
+ -89.4945068359375
221
+ ],
222
+ [
223
+ -51.648353576660156,
224
+ -74.24176025390625,
225
+ -69.23076629638672,
226
+ -97.67032623291016,
227
+ -92.9230728149414
228
+ ],
229
+ [
230
+ -52.96703338623047,
231
+ -78.54944610595703,
232
+ -70.54945373535156,
233
+ -102.5054931640625,
234
+ -95.9120864868164
235
+ ],
236
+ [
237
+ -54.28571319580078,
238
+ -82.85713958740234,
239
+ -73.0988998413086,
240
+ -107.25274658203125,
241
+ -98.54944610595703
242
+ ],
243
+ [
244
+ -55.604393005371094,
245
+ -86.989013671875,
246
+ -75.20878601074219,
247
+ -111.73626708984375,
248
+ -100.83515930175781
249
+ ]
250
+ ],
251
+ "max": [
252
+ [
253
+ 19.56043815612793,
254
+ 18.769229888916016,
255
+ 31.692306518554688,
256
+ 35.516483306884766,
257
+ 24.52747344970703
258
+ ],
259
+ [
260
+ 23.868131637573242,
261
+ 22.549449920654297,
262
+ 39.07692337036133,
263
+ 42.63736343383789,
264
+ 29.80219268798828
265
+ ],
266
+ [
267
+ 28.087913513183594,
268
+ 26.24176025390625,
269
+ 45.93406677246094,
270
+ 49.406593322753906,
271
+ 34.72527313232422
272
+ ],
273
+ [
274
+ 32.0439567565918,
275
+ 29.934066772460938,
276
+ 52.79121017456055,
277
+ 55.82417678833008,
278
+ 39.4725341796875
279
+ ],
280
+ [
281
+ 35.912086486816406,
282
+ 33.45055389404297,
283
+ 59.38461685180664,
284
+ 62.06593322753906,
285
+ 44.4835205078125
286
+ ],
287
+ [
288
+ 39.516483306884766,
289
+ 37.406593322753906,
290
+ 65.80220031738281,
291
+ 67.86812591552734,
292
+ 49.4945068359375
293
+ ],
294
+ [
295
+ 43.03296661376953,
296
+ 41.45054626464844,
297
+ 72.13186645507812,
298
+ 73.40658569335938,
299
+ 54.32966613769531
300
+ ],
301
+ [
302
+ 46.10988998413086,
303
+ 45.406593322753906,
304
+ 78.19779968261719,
305
+ 78.94505310058594,
306
+ 58.901092529296875
307
+ ],
308
+ [
309
+ 48.74725341796875,
310
+ 49.27472686767578,
311
+ 83.82417297363281,
312
+ 83.78021240234375,
313
+ 63.29669952392578
314
+ ],
315
+ [
316
+ 51.384613037109375,
317
+ 53.05494689941406,
318
+ 89.36264038085938,
319
+ 88.43955993652344,
320
+ 67.78022003173828
321
+ ],
322
+ [
323
+ 53.846153259277344,
324
+ 56.74725341796875,
325
+ 94.54945373535156,
326
+ 93.0988998413086,
327
+ 71.99999237060547
328
+ ],
329
+ [
330
+ 55.78022003173828,
331
+ 60.087913513183594,
332
+ 99.64835357666016,
333
+ 97.4945068359375,
334
+ 75.69229888916016
335
+ ],
336
+ [
337
+ 57.626373291015625,
338
+ 63.604393005371094,
339
+ 104.4835205078125,
340
+ 104.17582702636719,
341
+ 79.20878601074219
342
+ ],
343
+ [
344
+ 59.736263275146484,
345
+ 67.03296661376953,
346
+ 108.70329284667969,
347
+ 110.68132019042969,
348
+ 82.1098861694336
349
+ ],
350
+ [
351
+ 62.19780349731445,
352
+ 70.28571319580078,
353
+ 112.92308044433594,
354
+ 116.83515930175781,
355
+ 84.5714340209961
356
+ ],
357
+ [
358
+ 64.65933990478516,
359
+ 73.45054626464844,
360
+ 116.87911987304688,
361
+ 122.81318664550781,
362
+ 86.24176788330078
363
+ ]
364
+ ],
365
+ "mean": [
366
+ [
367
+ 0.09116898477077484,
368
+ -1.2357141971588135,
369
+ -0.8188950419425964,
370
+ -0.15813970565795898,
371
+ -0.4125640094280243
372
+ ],
373
+ [
374
+ 0.09542012959718704,
375
+ -1.2344847917556763,
376
+ -0.8190693259239197,
377
+ -0.15713702142238617,
378
+ -0.41158872842788696
379
+ ],
380
+ [
381
+ 0.09958027303218842,
382
+ -1.233577013015747,
383
+ -0.8192473649978638,
384
+ -0.15637809038162231,
385
+ -0.4105716645717621
386
+ ],
387
+ [
388
+ 0.10364068299531937,
389
+ -1.2329152822494507,
390
+ -0.8194178342819214,
391
+ -0.15579022467136383,
392
+ -0.4095292389392853
393
+ ],
394
+ [
395
+ 0.10757787525653839,
396
+ -1.2324373722076416,
397
+ -0.8195881843566895,
398
+ -0.15533040463924408,
399
+ -0.4084765911102295
400
+ ],
401
+ [
402
+ 0.11139658093452454,
403
+ -1.2320812940597534,
404
+ -0.8197605013847351,
405
+ -0.15492597222328186,
406
+ -0.40740954875946045
407
+ ],
408
+ [
409
+ 0.11513001471757889,
410
+ -1.2317777872085571,
411
+ -0.8199328780174255,
412
+ -0.15455682575702667,
413
+ -0.4063303768634796
414
+ ],
415
+ [
416
+ 0.11879564821720123,
417
+ -1.231497883796692,
418
+ -0.8201030492782593,
419
+ -0.15420255064964294,
420
+ -0.4052353501319885
421
+ ],
422
+ [
423
+ 0.12233330309391022,
424
+ -1.231207251548767,
425
+ -0.8202779293060303,
426
+ -0.15386062860488892,
427
+ -0.404136598110199
428
+ ],
429
+ [
430
+ 0.12573732435703278,
431
+ -1.23092520236969,
432
+ -0.8204483389854431,
433
+ -0.15353679656982422,
434
+ -0.403065025806427
435
+ ],
436
+ [
437
+ 0.12904979288578033,
438
+ -1.230648398399353,
439
+ -0.8206118941307068,
440
+ -0.1532251536846161,
441
+ -0.40204811096191406
442
+ ],
443
+ [
444
+ 0.13231366872787476,
445
+ -1.2303727865219116,
446
+ -0.8207806348800659,
447
+ -0.1529259979724884,
448
+ -0.40108177065849304
449
+ ],
450
+ [
451
+ 0.13556510210037231,
452
+ -1.2301011085510254,
453
+ -0.8209494352340698,
454
+ -0.1526346653699875,
455
+ -0.40014708042144775
456
+ ],
457
+ [
458
+ 0.13881613314151764,
459
+ -1.2298355102539062,
460
+ -0.8211175799369812,
461
+ -0.15234291553497314,
462
+ -0.3992435038089752
463
+ ],
464
+ [
465
+ 0.1420668512582779,
466
+ -1.2295620441436768,
467
+ -0.8212847709655762,
468
+ -0.15204375982284546,
469
+ -0.39837512373924255
470
+ ],
471
+ [
472
+ 0.1453183889389038,
473
+ -1.2292938232421875,
474
+ -0.82145094871521,
475
+ -0.1517469584941864,
476
+ -0.39751413464546204
477
+ ]
478
+ ],
479
+ "std": [
480
+ [
481
+ 2.8137104511260986,
482
+ 4.254403114318848,
483
+ 4.592617988586426,
484
+ 6.7580156326293945,
485
+ 3.4385712146759033
486
+ ],
487
+ [
488
+ 3.50972056388855,
489
+ 5.262273788452148,
490
+ 5.580240726470947,
491
+ 8.42597770690918,
492
+ 4.28427791595459
493
+ ],
494
+ [
495
+ 4.198278903961182,
496
+ 6.26309061050415,
497
+ 6.570157527923584,
498
+ 10.075050354003906,
499
+ 5.104696750640869
500
+ ],
501
+ [
502
+ 4.876944541931152,
503
+ 7.250848770141602,
504
+ 7.55081844329834,
505
+ 11.698025703430176,
506
+ 5.8973388671875
507
+ ],
508
+ [
509
+ 5.543993949890137,
510
+ 8.222739219665527,
511
+ 8.515984535217285,
512
+ 13.29118537902832,
513
+ 6.6630778312683105
514
+ ],
515
+ [
516
+ 6.198864459991455,
517
+ 9.176966667175293,
518
+ 9.462031364440918,
519
+ 14.852596282958984,
520
+ 7.403776168823242
521
+ ],
522
+ [
523
+ 6.841030120849609,
524
+ 10.112449645996094,
525
+ 10.387017250061035,
526
+ 16.381458282470703,
527
+ 8.121214866638184
528
+ ],
529
+ [
530
+ 7.470268249511719,
531
+ 11.028485298156738,
532
+ 11.289778709411621,
533
+ 17.87760353088379,
534
+ 8.816962242126465
535
+ ],
536
+ [
537
+ 8.086183547973633,
538
+ 11.924468040466309,
539
+ 12.169771194458008,
540
+ 19.341224670410156,
541
+ 9.492456436157227
542
+ ],
543
+ [
544
+ 8.688971519470215,
545
+ 12.80043888092041,
546
+ 13.026934623718262,
547
+ 20.772930145263672,
548
+ 10.148636817932129
549
+ ],
550
+ [
551
+ 9.278815269470215,
552
+ 13.656326293945312,
553
+ 13.861421585083008,
554
+ 22.173322677612305,
555
+ 10.786337852478027
556
+ ],
557
+ [
558
+ 9.855653762817383,
559
+ 14.49221134185791,
560
+ 14.673599243164062,
561
+ 23.543027877807617,
562
+ 11.406280517578125
563
+ ],
564
+ [
565
+ 10.420241355895996,
566
+ 15.308170318603516,
567
+ 15.464048385620117,
568
+ 24.88274574279785,
569
+ 12.009020805358887
570
+ ],
571
+ [
572
+ 10.972671508789062,
573
+ 16.104333877563477,
574
+ 16.233245849609375,
575
+ 26.19310760498047,
576
+ 12.594924926757812
577
+ ],
578
+ [
579
+ 11.513472557067871,
580
+ 16.881122589111328,
581
+ 16.981843948364258,
582
+ 27.475034713745117,
583
+ 13.16433048248291
584
+ ],
585
+ [
586
+ 12.04288101196289,
587
+ 17.63862419128418,
588
+ 17.710617065429688,
589
+ 28.72920799255371,
590
+ 13.717620849609375
591
+ ]
592
+ ],
593
+ "q01": [
594
+ [
595
+ -7.516483306884766,
596
+ -13.768792724609375,
597
+ -13.680882568359376,
598
+ -18.46154022216797,
599
+ -11.780227661132812
600
+ ],
601
+ [
602
+ -9.626373291015625,
603
+ -17.010990142822266,
604
+ -16.659339904785156,
605
+ -22.85714340209961,
606
+ -14.769226531982422
607
+ ],
608
+ [
609
+ -11.64835189819336,
610
+ -20.26374053955078,
611
+ -19.64834976196289,
612
+ -27.164836883544922,
613
+ -17.670333862304688
614
+ ],
615
+ [
616
+ -13.582418441772461,
617
+ -23.428570022583006,
618
+ -22.37362693786621,
619
+ -31.384615173339842,
620
+ -20.571426391601562
621
+ ],
622
+ [
623
+ -15.614945983886718,
624
+ -26.417583923339844,
625
+ -25.27472328186035,
626
+ -35.42857360839844,
627
+ -23.29669189453125
628
+ ],
629
+ [
630
+ -17.538461685180664,
631
+ -29.417142486572267,
632
+ -27.83472885131836,
633
+ -39.384613037109375,
634
+ -25.856701965332032
635
+ ],
636
+ [
637
+ -19.38461685180664,
638
+ -32.21977996826172,
639
+ -30.461536407470703,
640
+ -43.07692337036133,
641
+ -28.571426391601562
642
+ ],
643
+ [
644
+ -21.23076820373535,
645
+ -34.955609436035154,
646
+ -32.57142684936523,
647
+ -46.769229888916016,
648
+ -30.85714050292969
649
+ ],
650
+ [
651
+ -22.989011764526367,
652
+ -37.5824169921875,
653
+ -34.769229888916016,
654
+ -50.373626708984375,
655
+ -33.31868072509766
656
+ ],
657
+ [
658
+ -24.75780258178711,
659
+ -40.131866455078125,
660
+ -36.8896728515625,
661
+ -53.54901275634766,
662
+ -35.60438537597656
663
+ ],
664
+ [
665
+ -26.593406677246094,
666
+ -42.603955841064455,
667
+ -38.999562377929685,
668
+ -56.8896728515625,
669
+ -38.06593322753906
670
+ ],
671
+ [
672
+ -28.08791160583496,
673
+ -45.0549430847168,
674
+ -41.098899841308594,
675
+ -60.21977996826172,
676
+ -39.83472839355469
677
+ ],
678
+ [
679
+ -29.670330047607422,
680
+ -47.51648712158203,
681
+ -43.03296661376953,
682
+ -63.47252700805664,
683
+ -41.93406768798828
684
+ ],
685
+ [
686
+ -31.076923370361328,
687
+ -49.90066253662109,
688
+ -45.14285659790039,
689
+ -66.63735961914062,
690
+ -44.05450927734375
691
+ ],
692
+ [
693
+ -32.30769348144531,
694
+ -52.0,
695
+ -47.164833068847656,
696
+ -69.54900939941406,
697
+ -45.63692962646484
698
+ ],
699
+ [
700
+ -33.71428344726562,
701
+ -54.28571319580078,
702
+ -49.010986328125,
703
+ -72.17582702636719,
704
+ -47.120880126953125
705
+ ]
706
+ ],
707
+ "q99": [
708
+ [
709
+ 9.450550079345703,
710
+ 10.505495071411133,
711
+ 11.208789825439453,
712
+ 21.714283218383787,
713
+ 11.604400634765625
714
+ ],
715
+ [
716
+ 11.824174880981445,
717
+ 13.318679809570312,
718
+ 14.461536407470703,
719
+ 26.901100158691406,
720
+ 14.5054931640625
721
+ ],
722
+ [
723
+ 13.670330047607422,
724
+ 16.043960571289062,
725
+ 17.802200317382812,
726
+ 32.08790969848633,
727
+ 17.318680725097657
728
+ ],
729
+ [
730
+ 15.966595993042223,
731
+ 18.681320190429688,
732
+ 20.801759338379135,
733
+ 37.098896484375,
734
+ 19.956040344238282
735
+ ],
736
+ [
737
+ 18.153846740722656,
738
+ 21.318679809570312,
739
+ 23.878682708740463,
740
+ 41.93406677246094,
741
+ 22.5054931640625
742
+ ],
743
+ [
744
+ 20.439559936523438,
745
+ 23.868129959106444,
746
+ 26.857143630981444,
747
+ 46.593406677246094,
748
+ 25.054946899414062
749
+ ],
750
+ [
751
+ 22.461538314819336,
752
+ 26.24175666809082,
753
+ 29.846153717041016,
754
+ 51.35120925903343,
755
+ 27.25274658203125
756
+ ],
757
+ [
758
+ 24.48351692199707,
759
+ 28.625933990478746,
760
+ 32.659339904785156,
761
+ 55.82417343139648,
762
+ 29.450546264648438
763
+ ],
764
+ [
765
+ 26.417581787109373,
766
+ 30.901100158691406,
767
+ 35.39516601562523,
768
+ 60.21977996826172,
769
+ 31.648345947265625
770
+ ],
771
+ [
772
+ 28.527471771240233,
773
+ 33.10945281982445,
774
+ 37.93406677246094,
775
+ 64.43955993652344,
776
+ 33.856707763672105
777
+ ],
778
+ [
779
+ 30.296264266968002,
780
+ 35.20878982543945,
781
+ 40.49406921386742,
782
+ 68.65933990478516,
783
+ 35.79077636718773
784
+ ],
785
+ [
786
+ 32.21977996826172,
787
+ 37.153409118652576,
788
+ 43.21934280395531,
789
+ 72.53802307128929,
790
+ 37.724842224121325
791
+ ],
792
+ [
793
+ 33.98857269287132,
794
+ 39.17538269042992,
795
+ 45.67033386230469,
796
+ 76.57142639160156,
797
+ 39.82417297363281
798
+ ],
799
+ [
800
+ 35.64835021972656,
801
+ 40.92307662963867,
802
+ 48.406155700683826,
803
+ 80.61538696289062,
804
+ 41.41714874267601
805
+ ],
806
+ [
807
+ 37.23076629638672,
808
+ 42.5054931640625,
809
+ 50.60395629882836,
810
+ 84.49406921386742,
811
+ 43.42857360839844
812
+ ],
813
+ [
814
+ 38.90109634399414,
815
+ 44.087913513183594,
816
+ 52.80176269531273,
817
+ 88.26373291015625,
818
+ 45.27471923828125
819
+ ]
820
+ ]
821
+ }
822
+ }
823
+ }
824
+ }
experiment_cfg/final_model_config.json ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_type": "Gr00tN1d6",
3
+ "model_dtype": "bfloat16",
4
+ "model_name": "nvidia/Eagle-Block2A-2B-v2",
5
+ "backbone_model_type": "eagle",
6
+ "model_revision": null,
7
+ "tune_top_llm_layers": 4,
8
+ "backbone_embedding_dim": 2048,
9
+ "tune_llm": false,
10
+ "tune_visual": false,
11
+ "select_layer": 16,
12
+ "reproject_vision": false,
13
+ "use_flash_attention": true,
14
+ "load_bf16": true,
15
+ "collator_overwrite_image_inputs": false,
16
+ "eagle_collator": true,
17
+ "backbone_trainable_params_fp32": true,
18
+ "extra_augmentation_config": null,
19
+ "apply_sincos_state_encoding": true,
20
+ "use_relative_action": true,
21
+ "max_state_dim": 128,
22
+ "max_action_dim": 128,
23
+ "action_horizon": 50,
24
+ "hidden_size": 1024,
25
+ "input_embedding_dim": 1536,
26
+ "add_pos_embed": true,
27
+ "attn_dropout": 0.2,
28
+ "use_vlln": true,
29
+ "max_seq_len": 1024,
30
+ "use_alternate_vl_dit": true,
31
+ "attend_text_every_n_blocks": 2,
32
+ "diffusion_model_cfg": {
33
+ "attention_head_dim": 48,
34
+ "dropout": 0.2,
35
+ "final_dropout": true,
36
+ "interleave_self_attention": true,
37
+ "norm_type": "ada_norm",
38
+ "num_attention_heads": 32,
39
+ "num_layers": 32,
40
+ "output_dim": 1024,
41
+ "positional_embeddings": null
42
+ },
43
+ "num_inference_timesteps": 4,
44
+ "noise_beta_alpha": 1.5,
45
+ "noise_beta_beta": 1.0,
46
+ "noise_s": 0.999,
47
+ "num_timestep_buckets": 1000,
48
+ "tune_projector": true,
49
+ "tune_diffusion_model": true,
50
+ "tune_vlln": true,
51
+ "state_dropout_prob": 0.0,
52
+ "state_additive_noise_scale": 0.0,
53
+ "max_num_embodiments": 32
54
+ }
experiment_cfg/final_processor_config.json ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14487ee99ba8192d155f3a7f022bd4c1f7a7d2dda17d6bd9df10aa78a0cd7025
3
+ size 4990120184
model-00002-of-00002.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b2ff6f0a68fddfb8f2fa75a86e9f70072ecf1c36f58bb4421bce94e4bc98a80f
3
+ size 4823190320
model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0fb86946cf10594605f611aa66730ce237d923659829d00073402e27af5c730d
3
+ size 12960193762
processor_config.json ADDED
@@ -0,0 +1,455 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "processor_class": "Gr00tN1d6Processor",
3
+ "processor_kwargs": {
4
+ "modality_configs": {
5
+ "behavior_r1_pro": {
6
+ "video": {
7
+ "delta_indices": [
8
+ 0
9
+ ],
10
+ "modality_keys": [
11
+ "observation.images.rgb.head_256_256",
12
+ "observation.images.rgb.left_wrist_256_256",
13
+ "observation.images.rgb.right_wrist_256_256"
14
+ ],
15
+ "sin_cos_embedding_keys": null,
16
+ "mean_std_embedding_keys": null,
17
+ "action_configs": null
18
+ },
19
+ "state": {
20
+ "delta_indices": [
21
+ 0
22
+ ],
23
+ "modality_keys": [
24
+ "robot_pos",
25
+ "robot_ori_cos",
26
+ "robot_ori_sin",
27
+ "robot_2d_ori",
28
+ "robot_2d_ori_cos",
29
+ "robot_2d_ori_sin",
30
+ "robot_lin_vel",
31
+ "robot_ang_vel",
32
+ "arm_left_qpos",
33
+ "arm_left_qpos_sin",
34
+ "arm_left_qpos_cos",
35
+ "eef_left_pos",
36
+ "eef_left_quat",
37
+ "gripper_left_qpos",
38
+ "arm_right_qpos",
39
+ "arm_right_qpos_sin",
40
+ "arm_right_qpos_cos",
41
+ "eef_right_pos",
42
+ "eef_right_quat",
43
+ "gripper_right_qpos",
44
+ "trunk_qpos"
45
+ ],
46
+ "sin_cos_embedding_keys": null,
47
+ "mean_std_embedding_keys": null,
48
+ "action_configs": null
49
+ },
50
+ "action": {
51
+ "delta_indices": [
52
+ 0,
53
+ 1,
54
+ 2,
55
+ 3,
56
+ 4,
57
+ 5,
58
+ 6,
59
+ 7,
60
+ 8,
61
+ 9,
62
+ 10,
63
+ 11,
64
+ 12,
65
+ 13,
66
+ 14,
67
+ 15,
68
+ 16,
69
+ 17,
70
+ 18,
71
+ 19,
72
+ 20,
73
+ 21,
74
+ 22,
75
+ 23,
76
+ 24,
77
+ 25,
78
+ 26,
79
+ 27,
80
+ 28,
81
+ 29,
82
+ 30,
83
+ 31
84
+ ],
85
+ "modality_keys": [
86
+ "base",
87
+ "torso",
88
+ "left_arm",
89
+ "left_gripper",
90
+ "right_arm",
91
+ "right_gripper"
92
+ ],
93
+ "sin_cos_embedding_keys": null,
94
+ "mean_std_embedding_keys": null,
95
+ "action_configs": [
96
+ {
97
+ "rep": "ABSOLUTE",
98
+ "type": "NON_EEF",
99
+ "format": "DEFAULT",
100
+ "state_key": null
101
+ },
102
+ {
103
+ "rep": "RELATIVE",
104
+ "type": "NON_EEF",
105
+ "format": "DEFAULT",
106
+ "state_key": "trunk_qpos"
107
+ },
108
+ {
109
+ "rep": "RELATIVE",
110
+ "type": "NON_EEF",
111
+ "format": "DEFAULT",
112
+ "state_key": "arm_left_qpos"
113
+ },
114
+ {
115
+ "rep": "ABSOLUTE",
116
+ "type": "NON_EEF",
117
+ "format": "DEFAULT",
118
+ "state_key": null
119
+ },
120
+ {
121
+ "rep": "RELATIVE",
122
+ "type": "NON_EEF",
123
+ "format": "DEFAULT",
124
+ "state_key": "arm_right_qpos"
125
+ },
126
+ {
127
+ "rep": "ABSOLUTE",
128
+ "type": "NON_EEF",
129
+ "format": "DEFAULT",
130
+ "state_key": null
131
+ }
132
+ ]
133
+ },
134
+ "language": {
135
+ "delta_indices": [
136
+ 0
137
+ ],
138
+ "modality_keys": [
139
+ "annotation.human.coarse_action"
140
+ ],
141
+ "sin_cos_embedding_keys": null,
142
+ "mean_std_embedding_keys": null,
143
+ "action_configs": null
144
+ }
145
+ },
146
+ "gr1": {
147
+ "video": {
148
+ "delta_indices": [
149
+ 0
150
+ ],
151
+ "modality_keys": [
152
+ "ego_view_bg_crop_pad_res256_freq20"
153
+ ],
154
+ "sin_cos_embedding_keys": null,
155
+ "mean_std_embedding_keys": null,
156
+ "action_configs": null
157
+ },
158
+ "state": {
159
+ "delta_indices": [
160
+ 0
161
+ ],
162
+ "modality_keys": [
163
+ "left_arm",
164
+ "right_arm",
165
+ "left_hand",
166
+ "right_hand",
167
+ "waist"
168
+ ],
169
+ "sin_cos_embedding_keys": [
170
+ "left_arm",
171
+ "right_arm",
172
+ "left_hand",
173
+ "right_hand",
174
+ "waist"
175
+ ],
176
+ "mean_std_embedding_keys": null,
177
+ "action_configs": null
178
+ },
179
+ "action": {
180
+ "delta_indices": [
181
+ 0,
182
+ 1,
183
+ 2,
184
+ 3,
185
+ 4,
186
+ 5,
187
+ 6,
188
+ 7,
189
+ 8,
190
+ 9,
191
+ 10,
192
+ 11,
193
+ 12,
194
+ 13,
195
+ 14,
196
+ 15
197
+ ],
198
+ "modality_keys": [
199
+ "left_arm",
200
+ "right_arm",
201
+ "left_hand",
202
+ "right_hand",
203
+ "waist"
204
+ ],
205
+ "sin_cos_embedding_keys": null,
206
+ "mean_std_embedding_keys": null,
207
+ "action_configs": [
208
+ {
209
+ "rep": "RELATIVE",
210
+ "type": "NON_EEF",
211
+ "format": "DEFAULT",
212
+ "state_key": null
213
+ },
214
+ {
215
+ "rep": "RELATIVE",
216
+ "type": "NON_EEF",
217
+ "format": "DEFAULT",
218
+ "state_key": null
219
+ },
220
+ {
221
+ "rep": "RELATIVE",
222
+ "type": "NON_EEF",
223
+ "format": "DEFAULT",
224
+ "state_key": null
225
+ },
226
+ {
227
+ "rep": "RELATIVE",
228
+ "type": "NON_EEF",
229
+ "format": "DEFAULT",
230
+ "state_key": null
231
+ },
232
+ {
233
+ "rep": "ABSOLUTE",
234
+ "type": "NON_EEF",
235
+ "format": "DEFAULT",
236
+ "state_key": null
237
+ }
238
+ ]
239
+ },
240
+ "language": {
241
+ "delta_indices": [
242
+ 0
243
+ ],
244
+ "modality_keys": [
245
+ "task"
246
+ ],
247
+ "sin_cos_embedding_keys": null,
248
+ "mean_std_embedding_keys": null,
249
+ "action_configs": null
250
+ }
251
+ },
252
+ "robocasa_panda_omron": {
253
+ "video": {
254
+ "delta_indices": [
255
+ 0
256
+ ],
257
+ "modality_keys": [
258
+ "res256_image_side_0",
259
+ "res256_image_side_1",
260
+ "res256_image_wrist_0"
261
+ ],
262
+ "sin_cos_embedding_keys": null,
263
+ "mean_std_embedding_keys": null,
264
+ "action_configs": null
265
+ },
266
+ "state": {
267
+ "delta_indices": [
268
+ 0
269
+ ],
270
+ "modality_keys": [
271
+ "end_effector_position_relative",
272
+ "end_effector_rotation_relative",
273
+ "gripper_qpos",
274
+ "base_position",
275
+ "base_rotation"
276
+ ],
277
+ "sin_cos_embedding_keys": null,
278
+ "mean_std_embedding_keys": null,
279
+ "action_configs": null
280
+ },
281
+ "action": {
282
+ "delta_indices": [
283
+ 0,
284
+ 1,
285
+ 2,
286
+ 3,
287
+ 4,
288
+ 5,
289
+ 6,
290
+ 7,
291
+ 8,
292
+ 9,
293
+ 10,
294
+ 11,
295
+ 12,
296
+ 13,
297
+ 14,
298
+ 15
299
+ ],
300
+ "modality_keys": [
301
+ "end_effector_position",
302
+ "end_effector_rotation",
303
+ "gripper_close",
304
+ "base_motion",
305
+ "control_mode"
306
+ ],
307
+ "sin_cos_embedding_keys": null,
308
+ "mean_std_embedding_keys": null,
309
+ "action_configs": [
310
+ {
311
+ "rep": "ABSOLUTE",
312
+ "type": "NON_EEF",
313
+ "format": "DEFAULT",
314
+ "state_key": null
315
+ },
316
+ {
317
+ "rep": "ABSOLUTE",
318
+ "type": "NON_EEF",
319
+ "format": "DEFAULT",
320
+ "state_key": null
321
+ },
322
+ {
323
+ "rep": "ABSOLUTE",
324
+ "type": "NON_EEF",
325
+ "format": "DEFAULT",
326
+ "state_key": null
327
+ },
328
+ {
329
+ "rep": "ABSOLUTE",
330
+ "type": "NON_EEF",
331
+ "format": "DEFAULT",
332
+ "state_key": null
333
+ },
334
+ {
335
+ "rep": "ABSOLUTE",
336
+ "type": "NON_EEF",
337
+ "format": "DEFAULT",
338
+ "state_key": null
339
+ }
340
+ ]
341
+ },
342
+ "language": {
343
+ "delta_indices": [
344
+ 0
345
+ ],
346
+ "modality_keys": [
347
+ "annotation.human.action.task_description"
348
+ ],
349
+ "sin_cos_embedding_keys": null,
350
+ "mean_std_embedding_keys": null,
351
+ "action_configs": null
352
+ }
353
+ },
354
+ "new_embodiment": {
355
+ "video": {
356
+ "delta_indices": [
357
+ 0
358
+ ],
359
+ "modality_keys": [
360
+ "front",
361
+ "wrist"
362
+ ],
363
+ "sin_cos_embedding_keys": null,
364
+ "mean_std_embedding_keys": null,
365
+ "action_configs": null
366
+ },
367
+ "state": {
368
+ "delta_indices": [
369
+ 0
370
+ ],
371
+ "modality_keys": [
372
+ "single_arm",
373
+ "gripper"
374
+ ],
375
+ "sin_cos_embedding_keys": null,
376
+ "mean_std_embedding_keys": null,
377
+ "action_configs": null
378
+ },
379
+ "action": {
380
+ "delta_indices": [
381
+ 0,
382
+ 1,
383
+ 2,
384
+ 3,
385
+ 4,
386
+ 5,
387
+ 6,
388
+ 7,
389
+ 8,
390
+ 9,
391
+ 10,
392
+ 11,
393
+ 12,
394
+ 13,
395
+ 14,
396
+ 15
397
+ ],
398
+ "modality_keys": [
399
+ "single_arm",
400
+ "gripper"
401
+ ],
402
+ "sin_cos_embedding_keys": null,
403
+ "mean_std_embedding_keys": null,
404
+ "action_configs": [
405
+ {
406
+ "rep": "RELATIVE",
407
+ "type": "NON_EEF",
408
+ "format": "DEFAULT",
409
+ "state_key": null
410
+ },
411
+ {
412
+ "rep": "ABSOLUTE",
413
+ "type": "NON_EEF",
414
+ "format": "DEFAULT",
415
+ "state_key": null
416
+ }
417
+ ]
418
+ },
419
+ "language": {
420
+ "delta_indices": [
421
+ 0
422
+ ],
423
+ "modality_keys": [
424
+ "annotation.human.task_description"
425
+ ],
426
+ "sin_cos_embedding_keys": null,
427
+ "mean_std_embedding_keys": null,
428
+ "action_configs": null
429
+ }
430
+ }
431
+ },
432
+ "image_crop_size": null,
433
+ "image_target_size": null,
434
+ "use_albumentations": true,
435
+ "random_rotation_angle": null,
436
+ "color_jitter_params": {
437
+ "brightness": 0.3,
438
+ "contrast": 0.4,
439
+ "saturation": 0.5,
440
+ "hue": 0.08
441
+ },
442
+ "shortest_image_edge": 256,
443
+ "crop_fraction": 0.95,
444
+ "model_name": "nvidia/Eagle-Block2A-2B-v2",
445
+ "model_type": "eagle",
446
+ "formalize_language": true,
447
+ "max_state_dim": 128,
448
+ "max_action_dim": 128,
449
+ "max_action_horizon": 50,
450
+ "use_percentiles": false,
451
+ "clip_outliers": true,
452
+ "apply_sincos_state_encoding": true,
453
+ "use_relative_action": true
454
+ }
455
+ }
rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1f476c85815803955f3f0b961be1cdb4439238981e60db7010f58e8353145785
3
+ size 14645
scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c960d31406b345d9ea0d680e904d6145dbe0fc3b960fbf26ae378848e83ec25
3
+ size 1465
statistics.json ADDED
The diff for this file is too large to render. See raw diff
 
trainer_state.json ADDED
@@ -0,0 +1,3634 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0,
6
+ "eval_steps": 500,
7
+ "global_step": 6000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "grad_norm": 1.7000700235366821,
14
+ "learning_rate": 3e-06,
15
+ "loss": 1.1283,
16
+ "step": 10
17
+ },
18
+ {
19
+ "grad_norm": 0.8262935876846313,
20
+ "learning_rate": 6.333333333333334e-06,
21
+ "loss": 1.097,
22
+ "step": 20
23
+ },
24
+ {
25
+ "grad_norm": 0.7697754502296448,
26
+ "learning_rate": 9.666666666666667e-06,
27
+ "loss": 1.0692,
28
+ "step": 30
29
+ },
30
+ {
31
+ "grad_norm": 0.3612891733646393,
32
+ "learning_rate": 1.3000000000000001e-05,
33
+ "loss": 1.0448,
34
+ "step": 40
35
+ },
36
+ {
37
+ "grad_norm": 0.7433922290802002,
38
+ "learning_rate": 1.6333333333333335e-05,
39
+ "loss": 1.0697,
40
+ "step": 50
41
+ },
42
+ {
43
+ "grad_norm": 0.6279253959655762,
44
+ "learning_rate": 1.9666666666666666e-05,
45
+ "loss": 1.0638,
46
+ "step": 60
47
+ },
48
+ {
49
+ "grad_norm": 0.6754515767097473,
50
+ "learning_rate": 2.3000000000000003e-05,
51
+ "loss": 1.0625,
52
+ "step": 70
53
+ },
54
+ {
55
+ "grad_norm": 0.5152317881584167,
56
+ "learning_rate": 2.633333333333333e-05,
57
+ "loss": 1.068,
58
+ "step": 80
59
+ },
60
+ {
61
+ "grad_norm": 0.7765102386474609,
62
+ "learning_rate": 2.9666666666666672e-05,
63
+ "loss": 1.0458,
64
+ "step": 90
65
+ },
66
+ {
67
+ "grad_norm": 0.5475148558616638,
68
+ "learning_rate": 3.3e-05,
69
+ "loss": 1.0491,
70
+ "step": 100
71
+ },
72
+ {
73
+ "grad_norm": 0.9391897916793823,
74
+ "learning_rate": 3.633333333333333e-05,
75
+ "loss": 1.0308,
76
+ "step": 110
77
+ },
78
+ {
79
+ "grad_norm": 1.0357705354690552,
80
+ "learning_rate": 3.966666666666667e-05,
81
+ "loss": 0.9933,
82
+ "step": 120
83
+ },
84
+ {
85
+ "grad_norm": 0.7339250445365906,
86
+ "learning_rate": 4.3e-05,
87
+ "loss": 0.9393,
88
+ "step": 130
89
+ },
90
+ {
91
+ "grad_norm": 0.9334234595298767,
92
+ "learning_rate": 4.633333333333333e-05,
93
+ "loss": 0.9121,
94
+ "step": 140
95
+ },
96
+ {
97
+ "grad_norm": 4.375876426696777,
98
+ "learning_rate": 4.966666666666667e-05,
99
+ "loss": 0.85,
100
+ "step": 150
101
+ },
102
+ {
103
+ "grad_norm": 2.2090346813201904,
104
+ "learning_rate": 5.300000000000001e-05,
105
+ "loss": 0.7249,
106
+ "step": 160
107
+ },
108
+ {
109
+ "grad_norm": 1.9277912378311157,
110
+ "learning_rate": 5.633333333333334e-05,
111
+ "loss": 0.5533,
112
+ "step": 170
113
+ },
114
+ {
115
+ "grad_norm": 2.3054964542388916,
116
+ "learning_rate": 5.966666666666667e-05,
117
+ "loss": 0.4604,
118
+ "step": 180
119
+ },
120
+ {
121
+ "grad_norm": 1.3749431371688843,
122
+ "learning_rate": 6.3e-05,
123
+ "loss": 0.3421,
124
+ "step": 190
125
+ },
126
+ {
127
+ "grad_norm": 1.6281027793884277,
128
+ "learning_rate": 6.633333333333334e-05,
129
+ "loss": 0.2506,
130
+ "step": 200
131
+ },
132
+ {
133
+ "grad_norm": 1.410512089729309,
134
+ "learning_rate": 6.966666666666668e-05,
135
+ "loss": 0.189,
136
+ "step": 210
137
+ },
138
+ {
139
+ "grad_norm": 1.0307164192199707,
140
+ "learning_rate": 7.3e-05,
141
+ "loss": 0.1688,
142
+ "step": 220
143
+ },
144
+ {
145
+ "grad_norm": 1.0910820960998535,
146
+ "learning_rate": 7.633333333333334e-05,
147
+ "loss": 0.1543,
148
+ "step": 230
149
+ },
150
+ {
151
+ "grad_norm": 0.8725211024284363,
152
+ "learning_rate": 7.966666666666666e-05,
153
+ "loss": 0.1578,
154
+ "step": 240
155
+ },
156
+ {
157
+ "grad_norm": 1.1222732067108154,
158
+ "learning_rate": 8.3e-05,
159
+ "loss": 0.1432,
160
+ "step": 250
161
+ },
162
+ {
163
+ "grad_norm": 0.7612312436103821,
164
+ "learning_rate": 8.633333333333334e-05,
165
+ "loss": 0.13,
166
+ "step": 260
167
+ },
168
+ {
169
+ "grad_norm": 1.1903313398361206,
170
+ "learning_rate": 8.966666666666666e-05,
171
+ "loss": 0.148,
172
+ "step": 270
173
+ },
174
+ {
175
+ "grad_norm": 0.8089312314987183,
176
+ "learning_rate": 9.300000000000001e-05,
177
+ "loss": 0.1326,
178
+ "step": 280
179
+ },
180
+ {
181
+ "grad_norm": 0.9453487992286682,
182
+ "learning_rate": 9.633333333333335e-05,
183
+ "loss": 0.1467,
184
+ "step": 290
185
+ },
186
+ {
187
+ "grad_norm": 1.205849051475525,
188
+ "learning_rate": 9.966666666666667e-05,
189
+ "loss": 0.1273,
190
+ "step": 300
191
+ },
192
+ {
193
+ "grad_norm": 0.6407943964004517,
194
+ "learning_rate": 9.999938485971279e-05,
195
+ "loss": 0.1256,
196
+ "step": 310
197
+ },
198
+ {
199
+ "grad_norm": 0.696092963218689,
200
+ "learning_rate": 9.999725846827562e-05,
201
+ "loss": 0.113,
202
+ "step": 320
203
+ },
204
+ {
205
+ "grad_norm": 0.6457434892654419,
206
+ "learning_rate": 9.999361329594254e-05,
207
+ "loss": 0.1059,
208
+ "step": 330
209
+ },
210
+ {
211
+ "grad_norm": 0.769981324672699,
212
+ "learning_rate": 9.998844945344405e-05,
213
+ "loss": 0.102,
214
+ "step": 340
215
+ },
216
+ {
217
+ "grad_norm": 0.4546976387500763,
218
+ "learning_rate": 9.99817670976436e-05,
219
+ "loss": 0.0906,
220
+ "step": 350
221
+ },
222
+ {
223
+ "grad_norm": 0.7327650785446167,
224
+ "learning_rate": 9.997356643153303e-05,
225
+ "loss": 0.0935,
226
+ "step": 360
227
+ },
228
+ {
229
+ "grad_norm": 0.7533209919929504,
230
+ "learning_rate": 9.996384770422629e-05,
231
+ "loss": 0.1085,
232
+ "step": 370
233
+ },
234
+ {
235
+ "grad_norm": 0.6281214356422424,
236
+ "learning_rate": 9.995261121095194e-05,
237
+ "loss": 0.094,
238
+ "step": 380
239
+ },
240
+ {
241
+ "grad_norm": 0.45302775502204895,
242
+ "learning_rate": 9.993985729304408e-05,
243
+ "loss": 0.0973,
244
+ "step": 390
245
+ },
246
+ {
247
+ "grad_norm": 0.5847544074058533,
248
+ "learning_rate": 9.992558633793212e-05,
249
+ "loss": 0.0955,
250
+ "step": 400
251
+ },
252
+ {
253
+ "grad_norm": 0.5938292741775513,
254
+ "learning_rate": 9.990979877912891e-05,
255
+ "loss": 0.0927,
256
+ "step": 410
257
+ },
258
+ {
259
+ "grad_norm": 0.7282758355140686,
260
+ "learning_rate": 9.989249509621759e-05,
261
+ "loss": 0.0858,
262
+ "step": 420
263
+ },
264
+ {
265
+ "grad_norm": 0.6158447265625,
266
+ "learning_rate": 9.987367581483705e-05,
267
+ "loss": 0.0974,
268
+ "step": 430
269
+ },
270
+ {
271
+ "grad_norm": 0.4978296458721161,
272
+ "learning_rate": 9.985334150666592e-05,
273
+ "loss": 0.0946,
274
+ "step": 440
275
+ },
276
+ {
277
+ "grad_norm": 0.7747285962104797,
278
+ "learning_rate": 9.983149278940526e-05,
279
+ "loss": 0.0904,
280
+ "step": 450
281
+ },
282
+ {
283
+ "grad_norm": 0.5748892426490784,
284
+ "learning_rate": 9.980813032675974e-05,
285
+ "loss": 0.0938,
286
+ "step": 460
287
+ },
288
+ {
289
+ "grad_norm": 0.5600329637527466,
290
+ "learning_rate": 9.978325482841753e-05,
291
+ "loss": 0.0976,
292
+ "step": 470
293
+ },
294
+ {
295
+ "grad_norm": 0.4506227970123291,
296
+ "learning_rate": 9.975686705002867e-05,
297
+ "loss": 0.0875,
298
+ "step": 480
299
+ },
300
+ {
301
+ "grad_norm": 0.5217170715332031,
302
+ "learning_rate": 9.972896779318219e-05,
303
+ "loss": 0.085,
304
+ "step": 490
305
+ },
306
+ {
307
+ "grad_norm": 0.44347596168518066,
308
+ "learning_rate": 9.969955790538175e-05,
309
+ "loss": 0.0811,
310
+ "step": 500
311
+ },
312
+ {
313
+ "grad_norm": 0.7565658688545227,
314
+ "learning_rate": 9.966863828001982e-05,
315
+ "loss": 0.089,
316
+ "step": 510
317
+ },
318
+ {
319
+ "grad_norm": 0.4682891070842743,
320
+ "learning_rate": 9.963620985635065e-05,
321
+ "loss": 0.0801,
322
+ "step": 520
323
+ },
324
+ {
325
+ "grad_norm": 0.407950222492218,
326
+ "learning_rate": 9.960227361946164e-05,
327
+ "loss": 0.0855,
328
+ "step": 530
329
+ },
330
+ {
331
+ "grad_norm": 0.555381178855896,
332
+ "learning_rate": 9.95668306002435e-05,
333
+ "loss": 0.0769,
334
+ "step": 540
335
+ },
336
+ {
337
+ "grad_norm": 0.4560313820838928,
338
+ "learning_rate": 9.952988187535886e-05,
339
+ "loss": 0.0818,
340
+ "step": 550
341
+ },
342
+ {
343
+ "grad_norm": 0.46689373254776,
344
+ "learning_rate": 9.949142856720961e-05,
345
+ "loss": 0.0741,
346
+ "step": 560
347
+ },
348
+ {
349
+ "grad_norm": 0.5960952639579773,
350
+ "learning_rate": 9.945147184390278e-05,
351
+ "loss": 0.0744,
352
+ "step": 570
353
+ },
354
+ {
355
+ "grad_norm": 0.4881025552749634,
356
+ "learning_rate": 9.941001291921512e-05,
357
+ "loss": 0.0778,
358
+ "step": 580
359
+ },
360
+ {
361
+ "grad_norm": 0.44606727361679077,
362
+ "learning_rate": 9.936705305255612e-05,
363
+ "loss": 0.0805,
364
+ "step": 590
365
+ },
366
+ {
367
+ "grad_norm": 0.43473923206329346,
368
+ "learning_rate": 9.932259354892984e-05,
369
+ "loss": 0.0823,
370
+ "step": 600
371
+ },
372
+ {
373
+ "grad_norm": 0.39367181062698364,
374
+ "learning_rate": 9.927663575889521e-05,
375
+ "loss": 0.084,
376
+ "step": 610
377
+ },
378
+ {
379
+ "grad_norm": 0.47242826223373413,
380
+ "learning_rate": 9.922918107852504e-05,
381
+ "loss": 0.0729,
382
+ "step": 620
383
+ },
384
+ {
385
+ "grad_norm": 0.4946627616882324,
386
+ "learning_rate": 9.918023094936363e-05,
387
+ "loss": 0.0694,
388
+ "step": 630
389
+ },
390
+ {
391
+ "grad_norm": 0.4091627597808838,
392
+ "learning_rate": 9.912978685838294e-05,
393
+ "loss": 0.072,
394
+ "step": 640
395
+ },
396
+ {
397
+ "grad_norm": 0.2955147624015808,
398
+ "learning_rate": 9.90778503379374e-05,
399
+ "loss": 0.0704,
400
+ "step": 650
401
+ },
402
+ {
403
+ "grad_norm": 0.5645137429237366,
404
+ "learning_rate": 9.902442296571743e-05,
405
+ "loss": 0.0748,
406
+ "step": 660
407
+ },
408
+ {
409
+ "grad_norm": 0.5958395004272461,
410
+ "learning_rate": 9.896950636470147e-05,
411
+ "loss": 0.0883,
412
+ "step": 670
413
+ },
414
+ {
415
+ "grad_norm": 0.46314942836761475,
416
+ "learning_rate": 9.891310220310666e-05,
417
+ "loss": 0.077,
418
+ "step": 680
419
+ },
420
+ {
421
+ "grad_norm": 0.4550405442714691,
422
+ "learning_rate": 9.885521219433823e-05,
423
+ "loss": 0.0696,
424
+ "step": 690
425
+ },
426
+ {
427
+ "grad_norm": 0.36976656317710876,
428
+ "learning_rate": 9.879583809693738e-05,
429
+ "loss": 0.0719,
430
+ "step": 700
431
+ },
432
+ {
433
+ "grad_norm": 0.4276926815509796,
434
+ "learning_rate": 9.873498171452789e-05,
435
+ "loss": 0.0842,
436
+ "step": 710
437
+ },
438
+ {
439
+ "grad_norm": 0.6329816579818726,
440
+ "learning_rate": 9.867264489576135e-05,
441
+ "loss": 0.0767,
442
+ "step": 720
443
+ },
444
+ {
445
+ "grad_norm": 0.45453768968582153,
446
+ "learning_rate": 9.860882953426099e-05,
447
+ "loss": 0.0742,
448
+ "step": 730
449
+ },
450
+ {
451
+ "grad_norm": 0.4058501720428467,
452
+ "learning_rate": 9.854353756856412e-05,
453
+ "loss": 0.0643,
454
+ "step": 740
455
+ },
456
+ {
457
+ "grad_norm": 0.3971281349658966,
458
+ "learning_rate": 9.847677098206332e-05,
459
+ "loss": 0.0732,
460
+ "step": 750
461
+ },
462
+ {
463
+ "grad_norm": 0.44588446617126465,
464
+ "learning_rate": 9.840853180294608e-05,
465
+ "loss": 0.065,
466
+ "step": 760
467
+ },
468
+ {
469
+ "grad_norm": 0.4146692752838135,
470
+ "learning_rate": 9.833882210413332e-05,
471
+ "loss": 0.0704,
472
+ "step": 770
473
+ },
474
+ {
475
+ "grad_norm": 0.4290420413017273,
476
+ "learning_rate": 9.826764400321633e-05,
477
+ "loss": 0.0677,
478
+ "step": 780
479
+ },
480
+ {
481
+ "grad_norm": 0.4227924942970276,
482
+ "learning_rate": 9.819499966239243e-05,
483
+ "loss": 0.0703,
484
+ "step": 790
485
+ },
486
+ {
487
+ "grad_norm": 0.5033952593803406,
488
+ "learning_rate": 9.812089128839938e-05,
489
+ "loss": 0.0794,
490
+ "step": 800
491
+ },
492
+ {
493
+ "grad_norm": 0.3571436405181885,
494
+ "learning_rate": 9.804532113244828e-05,
495
+ "loss": 0.0717,
496
+ "step": 810
497
+ },
498
+ {
499
+ "grad_norm": 0.29655179381370544,
500
+ "learning_rate": 9.796829149015517e-05,
501
+ "loss": 0.0678,
502
+ "step": 820
503
+ },
504
+ {
505
+ "grad_norm": 0.43665337562561035,
506
+ "learning_rate": 9.788980470147132e-05,
507
+ "loss": 0.0668,
508
+ "step": 830
509
+ },
510
+ {
511
+ "grad_norm": 0.4904373586177826,
512
+ "learning_rate": 9.780986315061218e-05,
513
+ "loss": 0.0643,
514
+ "step": 840
515
+ },
516
+ {
517
+ "grad_norm": 0.3467118442058563,
518
+ "learning_rate": 9.772846926598491e-05,
519
+ "loss": 0.0752,
520
+ "step": 850
521
+ },
522
+ {
523
+ "grad_norm": 0.5342732667922974,
524
+ "learning_rate": 9.76456255201146e-05,
525
+ "loss": 0.0653,
526
+ "step": 860
527
+ },
528
+ {
529
+ "grad_norm": 0.5031263828277588,
530
+ "learning_rate": 9.756133442956923e-05,
531
+ "loss": 0.0638,
532
+ "step": 870
533
+ },
534
+ {
535
+ "grad_norm": 0.37514349818229675,
536
+ "learning_rate": 9.747559855488313e-05,
537
+ "loss": 0.063,
538
+ "step": 880
539
+ },
540
+ {
541
+ "grad_norm": 0.35820281505584717,
542
+ "learning_rate": 9.73884205004793e-05,
543
+ "loss": 0.0619,
544
+ "step": 890
545
+ },
546
+ {
547
+ "grad_norm": 0.5237784385681152,
548
+ "learning_rate": 9.729980291459019e-05,
549
+ "loss": 0.0694,
550
+ "step": 900
551
+ },
552
+ {
553
+ "grad_norm": 0.4646155536174774,
554
+ "learning_rate": 9.720974848917735e-05,
555
+ "loss": 0.0645,
556
+ "step": 910
557
+ },
558
+ {
559
+ "grad_norm": 0.4395473897457123,
560
+ "learning_rate": 9.711825995984957e-05,
561
+ "loss": 0.0687,
562
+ "step": 920
563
+ },
564
+ {
565
+ "grad_norm": 0.39834749698638916,
566
+ "learning_rate": 9.702534010577991e-05,
567
+ "loss": 0.0616,
568
+ "step": 930
569
+ },
570
+ {
571
+ "grad_norm": 0.43580004572868347,
572
+ "learning_rate": 9.693099174962103e-05,
573
+ "loss": 0.0602,
574
+ "step": 940
575
+ },
576
+ {
577
+ "grad_norm": 0.47095683217048645,
578
+ "learning_rate": 9.683521775741977e-05,
579
+ "loss": 0.0742,
580
+ "step": 950
581
+ },
582
+ {
583
+ "grad_norm": 0.4243863821029663,
584
+ "learning_rate": 9.673802103852979e-05,
585
+ "loss": 0.0569,
586
+ "step": 960
587
+ },
588
+ {
589
+ "grad_norm": 0.4031473398208618,
590
+ "learning_rate": 9.663940454552342e-05,
591
+ "loss": 0.0585,
592
+ "step": 970
593
+ },
594
+ {
595
+ "grad_norm": 0.47860366106033325,
596
+ "learning_rate": 9.65393712741018e-05,
597
+ "loss": 0.0634,
598
+ "step": 980
599
+ },
600
+ {
601
+ "grad_norm": 0.38193196058273315,
602
+ "learning_rate": 9.6437924263004e-05,
603
+ "loss": 0.0594,
604
+ "step": 990
605
+ },
606
+ {
607
+ "grad_norm": 0.4677029252052307,
608
+ "learning_rate": 9.63350665939146e-05,
609
+ "loss": 0.0617,
610
+ "step": 1000
611
+ },
612
+ {
613
+ "grad_norm": 0.3832956850528717,
614
+ "learning_rate": 9.623080139137023e-05,
615
+ "loss": 0.0616,
616
+ "step": 1010
617
+ },
618
+ {
619
+ "grad_norm": 0.36230915784835815,
620
+ "learning_rate": 9.612513182266447e-05,
621
+ "loss": 0.0615,
622
+ "step": 1020
623
+ },
624
+ {
625
+ "grad_norm": 0.4043103754520416,
626
+ "learning_rate": 9.601806109775179e-05,
627
+ "loss": 0.0664,
628
+ "step": 1030
629
+ },
630
+ {
631
+ "grad_norm": 0.43316778540611267,
632
+ "learning_rate": 9.590959246914995e-05,
633
+ "loss": 0.0663,
634
+ "step": 1040
635
+ },
636
+ {
637
+ "grad_norm": 0.32864686846733093,
638
+ "learning_rate": 9.579972923184122e-05,
639
+ "loss": 0.0603,
640
+ "step": 1050
641
+ },
642
+ {
643
+ "grad_norm": 0.3839098811149597,
644
+ "learning_rate": 9.568847472317232e-05,
645
+ "loss": 0.0642,
646
+ "step": 1060
647
+ },
648
+ {
649
+ "grad_norm": 0.28199896216392517,
650
+ "learning_rate": 9.557583232275303e-05,
651
+ "loss": 0.0604,
652
+ "step": 1070
653
+ },
654
+ {
655
+ "grad_norm": 0.3797306716442108,
656
+ "learning_rate": 9.546180545235344e-05,
657
+ "loss": 0.0601,
658
+ "step": 1080
659
+ },
660
+ {
661
+ "grad_norm": 0.2888736426830292,
662
+ "learning_rate": 9.534639757580013e-05,
663
+ "loss": 0.0495,
664
+ "step": 1090
665
+ },
666
+ {
667
+ "grad_norm": 0.2527846395969391,
668
+ "learning_rate": 9.522961219887092e-05,
669
+ "loss": 0.0542,
670
+ "step": 1100
671
+ },
672
+ {
673
+ "grad_norm": 0.26228803396224976,
674
+ "learning_rate": 9.511145286918828e-05,
675
+ "loss": 0.0571,
676
+ "step": 1110
677
+ },
678
+ {
679
+ "grad_norm": 0.5870105028152466,
680
+ "learning_rate": 9.499192317611167e-05,
681
+ "loss": 0.0649,
682
+ "step": 1120
683
+ },
684
+ {
685
+ "grad_norm": 0.39054936170578003,
686
+ "learning_rate": 9.487102675062851e-05,
687
+ "loss": 0.0637,
688
+ "step": 1130
689
+ },
690
+ {
691
+ "grad_norm": 0.3763654828071594,
692
+ "learning_rate": 9.474876726524374e-05,
693
+ "loss": 0.0612,
694
+ "step": 1140
695
+ },
696
+ {
697
+ "grad_norm": 0.2446664720773697,
698
+ "learning_rate": 9.462514843386845e-05,
699
+ "loss": 0.0572,
700
+ "step": 1150
701
+ },
702
+ {
703
+ "grad_norm": 0.29194676876068115,
704
+ "learning_rate": 9.450017401170689e-05,
705
+ "loss": 0.0516,
706
+ "step": 1160
707
+ },
708
+ {
709
+ "grad_norm": 0.51090008020401,
710
+ "learning_rate": 9.437384779514256e-05,
711
+ "loss": 0.0573,
712
+ "step": 1170
713
+ },
714
+ {
715
+ "grad_norm": 0.23249898850917816,
716
+ "learning_rate": 9.424617362162271e-05,
717
+ "loss": 0.0559,
718
+ "step": 1180
719
+ },
720
+ {
721
+ "grad_norm": 0.3077312409877777,
722
+ "learning_rate": 9.411715536954196e-05,
723
+ "loss": 0.054,
724
+ "step": 1190
725
+ },
726
+ {
727
+ "grad_norm": 0.42092403769493103,
728
+ "learning_rate": 9.39867969581243e-05,
729
+ "loss": 0.0574,
730
+ "step": 1200
731
+ },
732
+ {
733
+ "grad_norm": 0.45014843344688416,
734
+ "learning_rate": 9.385510234730415e-05,
735
+ "loss": 0.0518,
736
+ "step": 1210
737
+ },
738
+ {
739
+ "grad_norm": 0.45236918330192566,
740
+ "learning_rate": 9.372207553760603e-05,
741
+ "loss": 0.049,
742
+ "step": 1220
743
+ },
744
+ {
745
+ "grad_norm": 0.3155176341533661,
746
+ "learning_rate": 9.358772057002312e-05,
747
+ "loss": 0.0573,
748
+ "step": 1230
749
+ },
750
+ {
751
+ "grad_norm": 0.3163870573043823,
752
+ "learning_rate": 9.345204152589428e-05,
753
+ "loss": 0.0504,
754
+ "step": 1240
755
+ },
756
+ {
757
+ "grad_norm": 0.3211176097393036,
758
+ "learning_rate": 9.331504252678037e-05,
759
+ "loss": 0.0595,
760
+ "step": 1250
761
+ },
762
+ {
763
+ "grad_norm": 0.4440314769744873,
764
+ "learning_rate": 9.317672773433876e-05,
765
+ "loss": 0.0539,
766
+ "step": 1260
767
+ },
768
+ {
769
+ "grad_norm": 0.36051490902900696,
770
+ "learning_rate": 9.30371013501972e-05,
771
+ "loss": 0.06,
772
+ "step": 1270
773
+ },
774
+ {
775
+ "grad_norm": 0.43395301699638367,
776
+ "learning_rate": 9.289616761582587e-05,
777
+ "loss": 0.0592,
778
+ "step": 1280
779
+ },
780
+ {
781
+ "grad_norm": 0.40232110023498535,
782
+ "learning_rate": 9.275393081240882e-05,
783
+ "loss": 0.062,
784
+ "step": 1290
785
+ },
786
+ {
787
+ "grad_norm": 0.43771880865097046,
788
+ "learning_rate": 9.261039526071374e-05,
789
+ "loss": 0.0582,
790
+ "step": 1300
791
+ },
792
+ {
793
+ "grad_norm": 0.48542389273643494,
794
+ "learning_rate": 9.246556532096078e-05,
795
+ "loss": 0.0544,
796
+ "step": 1310
797
+ },
798
+ {
799
+ "grad_norm": 0.360439270734787,
800
+ "learning_rate": 9.231944539269009e-05,
801
+ "loss": 0.0562,
802
+ "step": 1320
803
+ },
804
+ {
805
+ "grad_norm": 0.3617100417613983,
806
+ "learning_rate": 9.217203991462815e-05,
807
+ "loss": 0.0521,
808
+ "step": 1330
809
+ },
810
+ {
811
+ "grad_norm": 0.3198077082633972,
812
+ "learning_rate": 9.202335336455296e-05,
813
+ "loss": 0.0524,
814
+ "step": 1340
815
+ },
816
+ {
817
+ "grad_norm": 0.36050671339035034,
818
+ "learning_rate": 9.187339025915802e-05,
819
+ "loss": 0.0496,
820
+ "step": 1350
821
+ },
822
+ {
823
+ "grad_norm": 0.309641033411026,
824
+ "learning_rate": 9.17221551539151e-05,
825
+ "loss": 0.0527,
826
+ "step": 1360
827
+ },
828
+ {
829
+ "grad_norm": 0.2965846359729767,
830
+ "learning_rate": 9.156965264293586e-05,
831
+ "loss": 0.0445,
832
+ "step": 1370
833
+ },
834
+ {
835
+ "grad_norm": 0.35571765899658203,
836
+ "learning_rate": 9.141588735883232e-05,
837
+ "loss": 0.0523,
838
+ "step": 1380
839
+ },
840
+ {
841
+ "grad_norm": 0.3687618374824524,
842
+ "learning_rate": 9.126086397257612e-05,
843
+ "loss": 0.0511,
844
+ "step": 1390
845
+ },
846
+ {
847
+ "grad_norm": 0.43302109837532043,
848
+ "learning_rate": 9.110458719335659e-05,
849
+ "loss": 0.0565,
850
+ "step": 1400
851
+ },
852
+ {
853
+ "grad_norm": 0.31573888659477234,
854
+ "learning_rate": 9.094706176843777e-05,
855
+ "loss": 0.0484,
856
+ "step": 1410
857
+ },
858
+ {
859
+ "grad_norm": 0.41654664278030396,
860
+ "learning_rate": 9.078829248301417e-05,
861
+ "loss": 0.0558,
862
+ "step": 1420
863
+ },
864
+ {
865
+ "grad_norm": 0.49469777941703796,
866
+ "learning_rate": 9.062828416006539e-05,
867
+ "loss": 0.0544,
868
+ "step": 1430
869
+ },
870
+ {
871
+ "grad_norm": 0.3401212990283966,
872
+ "learning_rate": 9.046704166020961e-05,
873
+ "loss": 0.0563,
874
+ "step": 1440
875
+ },
876
+ {
877
+ "grad_norm": 0.3731771409511566,
878
+ "learning_rate": 9.030456988155596e-05,
879
+ "loss": 0.0557,
880
+ "step": 1450
881
+ },
882
+ {
883
+ "grad_norm": 0.3299166262149811,
884
+ "learning_rate": 9.014087375955573e-05,
885
+ "loss": 0.0547,
886
+ "step": 1460
887
+ },
888
+ {
889
+ "grad_norm": 0.33593466877937317,
890
+ "learning_rate": 8.997595826685243e-05,
891
+ "loss": 0.0554,
892
+ "step": 1470
893
+ },
894
+ {
895
+ "grad_norm": 0.4328671097755432,
896
+ "learning_rate": 8.980982841313074e-05,
897
+ "loss": 0.0575,
898
+ "step": 1480
899
+ },
900
+ {
901
+ "grad_norm": 0.35234278440475464,
902
+ "learning_rate": 8.964248924496435e-05,
903
+ "loss": 0.0519,
904
+ "step": 1490
905
+ },
906
+ {
907
+ "grad_norm": 0.5080187320709229,
908
+ "learning_rate": 8.947394584566258e-05,
909
+ "loss": 0.0602,
910
+ "step": 1500
911
+ },
912
+ {
913
+ "grad_norm": 0.2649039626121521,
914
+ "learning_rate": 8.930420333511606e-05,
915
+ "loss": 0.053,
916
+ "step": 1510
917
+ },
918
+ {
919
+ "grad_norm": 0.42354223132133484,
920
+ "learning_rate": 8.913326686964117e-05,
921
+ "loss": 0.0501,
922
+ "step": 1520
923
+ },
924
+ {
925
+ "grad_norm": 0.4319063425064087,
926
+ "learning_rate": 8.89611416418234e-05,
927
+ "loss": 0.0538,
928
+ "step": 1530
929
+ },
930
+ {
931
+ "grad_norm": 0.4024607539176941,
932
+ "learning_rate": 8.878783288035957e-05,
933
+ "loss": 0.0616,
934
+ "step": 1540
935
+ },
936
+ {
937
+ "grad_norm": 0.34141480922698975,
938
+ "learning_rate": 8.86133458498991e-05,
939
+ "loss": 0.0514,
940
+ "step": 1550
941
+ },
942
+ {
943
+ "grad_norm": 0.3257942795753479,
944
+ "learning_rate": 8.843768585088393e-05,
945
+ "loss": 0.0569,
946
+ "step": 1560
947
+ },
948
+ {
949
+ "grad_norm": 0.22518017888069153,
950
+ "learning_rate": 8.82608582193877e-05,
951
+ "loss": 0.048,
952
+ "step": 1570
953
+ },
954
+ {
955
+ "grad_norm": 0.358802855014801,
956
+ "learning_rate": 8.80828683269535e-05,
957
+ "loss": 0.0516,
958
+ "step": 1580
959
+ },
960
+ {
961
+ "grad_norm": 0.2343483865261078,
962
+ "learning_rate": 8.790372158043074e-05,
963
+ "loss": 0.0494,
964
+ "step": 1590
965
+ },
966
+ {
967
+ "grad_norm": 0.37448617815971375,
968
+ "learning_rate": 8.772342342181095e-05,
969
+ "loss": 0.0477,
970
+ "step": 1600
971
+ },
972
+ {
973
+ "grad_norm": 0.36448949575424194,
974
+ "learning_rate": 8.75419793280624e-05,
975
+ "loss": 0.0499,
976
+ "step": 1610
977
+ },
978
+ {
979
+ "grad_norm": 0.2531074583530426,
980
+ "learning_rate": 8.735939481096378e-05,
981
+ "loss": 0.0425,
982
+ "step": 1620
983
+ },
984
+ {
985
+ "grad_norm": 0.32438674569129944,
986
+ "learning_rate": 8.717567541693673e-05,
987
+ "loss": 0.0433,
988
+ "step": 1630
989
+ },
990
+ {
991
+ "grad_norm": 0.3196386992931366,
992
+ "learning_rate": 8.699082672687734e-05,
993
+ "loss": 0.051,
994
+ "step": 1640
995
+ },
996
+ {
997
+ "grad_norm": 0.4035435616970062,
998
+ "learning_rate": 8.680485435598673e-05,
999
+ "loss": 0.0541,
1000
+ "step": 1650
1001
+ },
1002
+ {
1003
+ "grad_norm": 0.2992634177207947,
1004
+ "learning_rate": 8.661776395360029e-05,
1005
+ "loss": 0.0448,
1006
+ "step": 1660
1007
+ },
1008
+ {
1009
+ "grad_norm": 0.3280009329319,
1010
+ "learning_rate": 8.642956120301626e-05,
1011
+ "loss": 0.047,
1012
+ "step": 1670
1013
+ },
1014
+ {
1015
+ "grad_norm": 0.25291767716407776,
1016
+ "learning_rate": 8.624025182132292e-05,
1017
+ "loss": 0.0473,
1018
+ "step": 1680
1019
+ },
1020
+ {
1021
+ "grad_norm": 0.28412434458732605,
1022
+ "learning_rate": 8.604984155922506e-05,
1023
+ "loss": 0.0518,
1024
+ "step": 1690
1025
+ },
1026
+ {
1027
+ "grad_norm": 0.3839319944381714,
1028
+ "learning_rate": 8.585833620086918e-05,
1029
+ "loss": 0.0467,
1030
+ "step": 1700
1031
+ },
1032
+ {
1033
+ "grad_norm": 0.2821938991546631,
1034
+ "learning_rate": 8.566574156366784e-05,
1035
+ "loss": 0.0486,
1036
+ "step": 1710
1037
+ },
1038
+ {
1039
+ "grad_norm": 0.33787041902542114,
1040
+ "learning_rate": 8.547206349812298e-05,
1041
+ "loss": 0.0483,
1042
+ "step": 1720
1043
+ },
1044
+ {
1045
+ "grad_norm": 0.39908427000045776,
1046
+ "learning_rate": 8.527730788764805e-05,
1047
+ "loss": 0.0487,
1048
+ "step": 1730
1049
+ },
1050
+ {
1051
+ "grad_norm": 0.35582125186920166,
1052
+ "learning_rate": 8.508148064838948e-05,
1053
+ "loss": 0.045,
1054
+ "step": 1740
1055
+ },
1056
+ {
1057
+ "grad_norm": 0.42031681537628174,
1058
+ "learning_rate": 8.488458772904684e-05,
1059
+ "loss": 0.05,
1060
+ "step": 1750
1061
+ },
1062
+ {
1063
+ "grad_norm": 0.3604871928691864,
1064
+ "learning_rate": 8.468663511069217e-05,
1065
+ "loss": 0.0508,
1066
+ "step": 1760
1067
+ },
1068
+ {
1069
+ "grad_norm": 0.3763245940208435,
1070
+ "learning_rate": 8.448762880658825e-05,
1071
+ "loss": 0.0523,
1072
+ "step": 1770
1073
+ },
1074
+ {
1075
+ "grad_norm": 0.2665746212005615,
1076
+ "learning_rate": 8.428757486200603e-05,
1077
+ "loss": 0.0466,
1078
+ "step": 1780
1079
+ },
1080
+ {
1081
+ "grad_norm": 0.30607786774635315,
1082
+ "learning_rate": 8.40864793540409e-05,
1083
+ "loss": 0.0488,
1084
+ "step": 1790
1085
+ },
1086
+ {
1087
+ "grad_norm": 0.42583027482032776,
1088
+ "learning_rate": 8.388434839142813e-05,
1089
+ "loss": 0.0446,
1090
+ "step": 1800
1091
+ },
1092
+ {
1093
+ "grad_norm": 0.28605785965919495,
1094
+ "learning_rate": 8.368118811435726e-05,
1095
+ "loss": 0.0463,
1096
+ "step": 1810
1097
+ },
1098
+ {
1099
+ "grad_norm": 0.32007771730422974,
1100
+ "learning_rate": 8.347700469428564e-05,
1101
+ "loss": 0.0464,
1102
+ "step": 1820
1103
+ },
1104
+ {
1105
+ "grad_norm": 0.4149059057235718,
1106
+ "learning_rate": 8.327180433375091e-05,
1107
+ "loss": 0.0469,
1108
+ "step": 1830
1109
+ },
1110
+ {
1111
+ "grad_norm": 0.3714878559112549,
1112
+ "learning_rate": 8.306559326618259e-05,
1113
+ "loss": 0.0468,
1114
+ "step": 1840
1115
+ },
1116
+ {
1117
+ "grad_norm": 0.33420297503471375,
1118
+ "learning_rate": 8.285837775571276e-05,
1119
+ "loss": 0.0435,
1120
+ "step": 1850
1121
+ },
1122
+ {
1123
+ "grad_norm": 0.22721770405769348,
1124
+ "learning_rate": 8.265016409698573e-05,
1125
+ "loss": 0.047,
1126
+ "step": 1860
1127
+ },
1128
+ {
1129
+ "grad_norm": 0.3329220116138458,
1130
+ "learning_rate": 8.244095861496686e-05,
1131
+ "loss": 0.0504,
1132
+ "step": 1870
1133
+ },
1134
+ {
1135
+ "grad_norm": 0.479600727558136,
1136
+ "learning_rate": 8.223076766475035e-05,
1137
+ "loss": 0.042,
1138
+ "step": 1880
1139
+ },
1140
+ {
1141
+ "grad_norm": 0.4081374704837799,
1142
+ "learning_rate": 8.201959763136633e-05,
1143
+ "loss": 0.0433,
1144
+ "step": 1890
1145
+ },
1146
+ {
1147
+ "grad_norm": 0.2546713054180145,
1148
+ "learning_rate": 8.180745492958674e-05,
1149
+ "loss": 0.0533,
1150
+ "step": 1900
1151
+ },
1152
+ {
1153
+ "grad_norm": 0.25530946254730225,
1154
+ "learning_rate": 8.159434600373061e-05,
1155
+ "loss": 0.0486,
1156
+ "step": 1910
1157
+ },
1158
+ {
1159
+ "grad_norm": 0.370567262172699,
1160
+ "learning_rate": 8.138027732746818e-05,
1161
+ "loss": 0.0452,
1162
+ "step": 1920
1163
+ },
1164
+ {
1165
+ "grad_norm": 0.34710201621055603,
1166
+ "learning_rate": 8.116525540362434e-05,
1167
+ "loss": 0.0463,
1168
+ "step": 1930
1169
+ },
1170
+ {
1171
+ "grad_norm": 0.30122607946395874,
1172
+ "learning_rate": 8.094928676398101e-05,
1173
+ "loss": 0.0455,
1174
+ "step": 1940
1175
+ },
1176
+ {
1177
+ "grad_norm": 0.24021995067596436,
1178
+ "learning_rate": 8.073237796907882e-05,
1179
+ "loss": 0.0472,
1180
+ "step": 1950
1181
+ },
1182
+ {
1183
+ "grad_norm": 0.3200050890445709,
1184
+ "learning_rate": 8.051453560801772e-05,
1185
+ "loss": 0.0413,
1186
+ "step": 1960
1187
+ },
1188
+ {
1189
+ "grad_norm": 0.2531821131706238,
1190
+ "learning_rate": 8.029576629825687e-05,
1191
+ "loss": 0.0489,
1192
+ "step": 1970
1193
+ },
1194
+ {
1195
+ "grad_norm": 0.3085343539714813,
1196
+ "learning_rate": 8.007607668541362e-05,
1197
+ "loss": 0.0532,
1198
+ "step": 1980
1199
+ },
1200
+ {
1201
+ "grad_norm": 0.3176669776439667,
1202
+ "learning_rate": 7.985547344306161e-05,
1203
+ "loss": 0.0438,
1204
+ "step": 1990
1205
+ },
1206
+ {
1207
+ "grad_norm": 0.3424891531467438,
1208
+ "learning_rate": 7.963396327252812e-05,
1209
+ "loss": 0.0417,
1210
+ "step": 2000
1211
+ },
1212
+ {
1213
+ "grad_norm": 0.2856842279434204,
1214
+ "learning_rate": 7.941155290269038e-05,
1215
+ "loss": 0.0398,
1216
+ "step": 2010
1217
+ },
1218
+ {
1219
+ "grad_norm": 0.2625652551651001,
1220
+ "learning_rate": 7.918824908977123e-05,
1221
+ "loss": 0.0404,
1222
+ "step": 2020
1223
+ },
1224
+ {
1225
+ "grad_norm": 0.33125776052474976,
1226
+ "learning_rate": 7.896405861713394e-05,
1227
+ "loss": 0.0455,
1228
+ "step": 2030
1229
+ },
1230
+ {
1231
+ "grad_norm": 0.36468076705932617,
1232
+ "learning_rate": 7.873898829507606e-05,
1233
+ "loss": 0.0499,
1234
+ "step": 2040
1235
+ },
1236
+ {
1237
+ "grad_norm": 0.3898219168186188,
1238
+ "learning_rate": 7.851304496062254e-05,
1239
+ "loss": 0.0433,
1240
+ "step": 2050
1241
+ },
1242
+ {
1243
+ "grad_norm": 0.3178541958332062,
1244
+ "learning_rate": 7.828623547731818e-05,
1245
+ "loss": 0.0375,
1246
+ "step": 2060
1247
+ },
1248
+ {
1249
+ "grad_norm": 0.2662155032157898,
1250
+ "learning_rate": 7.80585667350189e-05,
1251
+ "loss": 0.0431,
1252
+ "step": 2070
1253
+ },
1254
+ {
1255
+ "grad_norm": 0.19700299203395844,
1256
+ "learning_rate": 7.783004564968263e-05,
1257
+ "loss": 0.0359,
1258
+ "step": 2080
1259
+ },
1260
+ {
1261
+ "grad_norm": 0.2499379962682724,
1262
+ "learning_rate": 7.760067916315921e-05,
1263
+ "loss": 0.035,
1264
+ "step": 2090
1265
+ },
1266
+ {
1267
+ "grad_norm": 0.18432468175888062,
1268
+ "learning_rate": 7.737047424297941e-05,
1269
+ "loss": 0.0352,
1270
+ "step": 2100
1271
+ },
1272
+ {
1273
+ "grad_norm": 0.3939158022403717,
1274
+ "learning_rate": 7.713943788214337e-05,
1275
+ "loss": 0.0394,
1276
+ "step": 2110
1277
+ },
1278
+ {
1279
+ "grad_norm": 0.430083692073822,
1280
+ "learning_rate": 7.690757709890812e-05,
1281
+ "loss": 0.0378,
1282
+ "step": 2120
1283
+ },
1284
+ {
1285
+ "grad_norm": 0.23813720047473907,
1286
+ "learning_rate": 7.66748989365744e-05,
1287
+ "loss": 0.041,
1288
+ "step": 2130
1289
+ },
1290
+ {
1291
+ "grad_norm": 0.30467575788497925,
1292
+ "learning_rate": 7.644141046327271e-05,
1293
+ "loss": 0.0469,
1294
+ "step": 2140
1295
+ },
1296
+ {
1297
+ "grad_norm": 0.3030410408973694,
1298
+ "learning_rate": 7.620711877174866e-05,
1299
+ "loss": 0.0424,
1300
+ "step": 2150
1301
+ },
1302
+ {
1303
+ "grad_norm": 0.23633800446987152,
1304
+ "learning_rate": 7.597203097914732e-05,
1305
+ "loss": 0.0393,
1306
+ "step": 2160
1307
+ },
1308
+ {
1309
+ "grad_norm": 0.31460145115852356,
1310
+ "learning_rate": 7.573615422679726e-05,
1311
+ "loss": 0.0362,
1312
+ "step": 2170
1313
+ },
1314
+ {
1315
+ "grad_norm": 0.4296618700027466,
1316
+ "learning_rate": 7.549949567999345e-05,
1317
+ "loss": 0.0437,
1318
+ "step": 2180
1319
+ },
1320
+ {
1321
+ "grad_norm": 0.2874586582183838,
1322
+ "learning_rate": 7.526206252777968e-05,
1323
+ "loss": 0.0341,
1324
+ "step": 2190
1325
+ },
1326
+ {
1327
+ "grad_norm": 0.36963751912117004,
1328
+ "learning_rate": 7.50238619827301e-05,
1329
+ "loss": 0.0387,
1330
+ "step": 2200
1331
+ },
1332
+ {
1333
+ "grad_norm": 0.38531920313835144,
1334
+ "learning_rate": 7.478490128073022e-05,
1335
+ "loss": 0.0367,
1336
+ "step": 2210
1337
+ },
1338
+ {
1339
+ "grad_norm": 0.22951769828796387,
1340
+ "learning_rate": 7.454518768075704e-05,
1341
+ "loss": 0.0385,
1342
+ "step": 2220
1343
+ },
1344
+ {
1345
+ "grad_norm": 0.32662105560302734,
1346
+ "learning_rate": 7.430472846465856e-05,
1347
+ "loss": 0.0453,
1348
+ "step": 2230
1349
+ },
1350
+ {
1351
+ "grad_norm": 0.22467322647571564,
1352
+ "learning_rate": 7.406353093693253e-05,
1353
+ "loss": 0.0361,
1354
+ "step": 2240
1355
+ },
1356
+ {
1357
+ "grad_norm": 0.20656687021255493,
1358
+ "learning_rate": 7.382160242450469e-05,
1359
+ "loss": 0.0408,
1360
+ "step": 2250
1361
+ },
1362
+ {
1363
+ "grad_norm": 0.340875506401062,
1364
+ "learning_rate": 7.357895027650598e-05,
1365
+ "loss": 0.0388,
1366
+ "step": 2260
1367
+ },
1368
+ {
1369
+ "grad_norm": 0.18724791705608368,
1370
+ "learning_rate": 7.333558186404958e-05,
1371
+ "loss": 0.0375,
1372
+ "step": 2270
1373
+ },
1374
+ {
1375
+ "grad_norm": 0.21643663942813873,
1376
+ "learning_rate": 7.309150458000668e-05,
1377
+ "loss": 0.0427,
1378
+ "step": 2280
1379
+ },
1380
+ {
1381
+ "grad_norm": 0.22064824402332306,
1382
+ "learning_rate": 7.284672583878219e-05,
1383
+ "loss": 0.0375,
1384
+ "step": 2290
1385
+ },
1386
+ {
1387
+ "grad_norm": 0.25080978870391846,
1388
+ "learning_rate": 7.260125307608929e-05,
1389
+ "loss": 0.0337,
1390
+ "step": 2300
1391
+ },
1392
+ {
1393
+ "grad_norm": 0.3313097357749939,
1394
+ "learning_rate": 7.235509374872373e-05,
1395
+ "loss": 0.0431,
1396
+ "step": 2310
1397
+ },
1398
+ {
1399
+ "grad_norm": 0.28288960456848145,
1400
+ "learning_rate": 7.210825533433719e-05,
1401
+ "loss": 0.0375,
1402
+ "step": 2320
1403
+ },
1404
+ {
1405
+ "grad_norm": 0.3031257390975952,
1406
+ "learning_rate": 7.186074533121013e-05,
1407
+ "loss": 0.048,
1408
+ "step": 2330
1409
+ },
1410
+ {
1411
+ "grad_norm": 0.2575356960296631,
1412
+ "learning_rate": 7.161257125802413e-05,
1413
+ "loss": 0.0375,
1414
+ "step": 2340
1415
+ },
1416
+ {
1417
+ "grad_norm": 0.264274001121521,
1418
+ "learning_rate": 7.136374065363334e-05,
1419
+ "loss": 0.0399,
1420
+ "step": 2350
1421
+ },
1422
+ {
1423
+ "grad_norm": 0.3352709412574768,
1424
+ "learning_rate": 7.11142610768356e-05,
1425
+ "loss": 0.0345,
1426
+ "step": 2360
1427
+ },
1428
+ {
1429
+ "grad_norm": 0.2115754336118698,
1430
+ "learning_rate": 7.086414010614276e-05,
1431
+ "loss": 0.0374,
1432
+ "step": 2370
1433
+ },
1434
+ {
1435
+ "grad_norm": 0.2517990171909332,
1436
+ "learning_rate": 7.061338533955043e-05,
1437
+ "loss": 0.0324,
1438
+ "step": 2380
1439
+ },
1440
+ {
1441
+ "grad_norm": 0.24857991933822632,
1442
+ "learning_rate": 7.036200439430725e-05,
1443
+ "loss": 0.0391,
1444
+ "step": 2390
1445
+ },
1446
+ {
1447
+ "grad_norm": 0.23163624107837677,
1448
+ "learning_rate": 7.01100049066835e-05,
1449
+ "loss": 0.0389,
1450
+ "step": 2400
1451
+ },
1452
+ {
1453
+ "grad_norm": 0.33151012659072876,
1454
+ "learning_rate": 6.985739453173903e-05,
1455
+ "loss": 0.033,
1456
+ "step": 2410
1457
+ },
1458
+ {
1459
+ "grad_norm": 0.26381349563598633,
1460
+ "learning_rate": 6.960418094309085e-05,
1461
+ "loss": 0.0341,
1462
+ "step": 2420
1463
+ },
1464
+ {
1465
+ "grad_norm": 0.30055221915245056,
1466
+ "learning_rate": 6.93503718326799e-05,
1467
+ "loss": 0.0336,
1468
+ "step": 2430
1469
+ },
1470
+ {
1471
+ "grad_norm": 0.25649207830429077,
1472
+ "learning_rate": 6.909597491053751e-05,
1473
+ "loss": 0.0389,
1474
+ "step": 2440
1475
+ },
1476
+ {
1477
+ "grad_norm": 0.2667904794216156,
1478
+ "learning_rate": 6.884099790455113e-05,
1479
+ "loss": 0.0321,
1480
+ "step": 2450
1481
+ },
1482
+ {
1483
+ "grad_norm": 0.38824063539505005,
1484
+ "learning_rate": 6.858544856022952e-05,
1485
+ "loss": 0.0393,
1486
+ "step": 2460
1487
+ },
1488
+ {
1489
+ "grad_norm": 0.2738566994667053,
1490
+ "learning_rate": 6.83293346404676e-05,
1491
+ "loss": 0.0377,
1492
+ "step": 2470
1493
+ },
1494
+ {
1495
+ "grad_norm": 0.2428845465183258,
1496
+ "learning_rate": 6.80726639253105e-05,
1497
+ "loss": 0.0332,
1498
+ "step": 2480
1499
+ },
1500
+ {
1501
+ "grad_norm": 0.27254635095596313,
1502
+ "learning_rate": 6.781544421171732e-05,
1503
+ "loss": 0.0311,
1504
+ "step": 2490
1505
+ },
1506
+ {
1507
+ "grad_norm": 0.3009340465068817,
1508
+ "learning_rate": 6.755768331332424e-05,
1509
+ "loss": 0.0361,
1510
+ "step": 2500
1511
+ },
1512
+ {
1513
+ "grad_norm": 0.27296382188796997,
1514
+ "learning_rate": 6.729938906020713e-05,
1515
+ "loss": 0.0388,
1516
+ "step": 2510
1517
+ },
1518
+ {
1519
+ "grad_norm": 0.19399593770503998,
1520
+ "learning_rate": 6.704056929864376e-05,
1521
+ "loss": 0.0348,
1522
+ "step": 2520
1523
+ },
1524
+ {
1525
+ "grad_norm": 0.3452855050563812,
1526
+ "learning_rate": 6.67812318908754e-05,
1527
+ "loss": 0.0313,
1528
+ "step": 2530
1529
+ },
1530
+ {
1531
+ "grad_norm": 0.2279820591211319,
1532
+ "learning_rate": 6.6521384714868e-05,
1533
+ "loss": 0.0372,
1534
+ "step": 2540
1535
+ },
1536
+ {
1537
+ "grad_norm": 0.23924601078033447,
1538
+ "learning_rate": 6.626103566407295e-05,
1539
+ "loss": 0.0421,
1540
+ "step": 2550
1541
+ },
1542
+ {
1543
+ "grad_norm": 0.26451361179351807,
1544
+ "learning_rate": 6.600019264718713e-05,
1545
+ "loss": 0.0364,
1546
+ "step": 2560
1547
+ },
1548
+ {
1549
+ "grad_norm": 0.24230974912643433,
1550
+ "learning_rate": 6.573886358791285e-05,
1551
+ "loss": 0.0356,
1552
+ "step": 2570
1553
+ },
1554
+ {
1555
+ "grad_norm": 0.21219466626644135,
1556
+ "learning_rate": 6.547705642471703e-05,
1557
+ "loss": 0.0357,
1558
+ "step": 2580
1559
+ },
1560
+ {
1561
+ "grad_norm": 0.21990950405597687,
1562
+ "learning_rate": 6.521477911059008e-05,
1563
+ "loss": 0.0369,
1564
+ "step": 2590
1565
+ },
1566
+ {
1567
+ "grad_norm": 0.25543391704559326,
1568
+ "learning_rate": 6.495203961280434e-05,
1569
+ "loss": 0.0332,
1570
+ "step": 2600
1571
+ },
1572
+ {
1573
+ "grad_norm": 0.23028869926929474,
1574
+ "learning_rate": 6.468884591267204e-05,
1575
+ "loss": 0.0317,
1576
+ "step": 2610
1577
+ },
1578
+ {
1579
+ "grad_norm": 0.25860467553138733,
1580
+ "learning_rate": 6.44252060053028e-05,
1581
+ "loss": 0.0341,
1582
+ "step": 2620
1583
+ },
1584
+ {
1585
+ "grad_norm": 0.23476362228393555,
1586
+ "learning_rate": 6.416112789936086e-05,
1587
+ "loss": 0.0304,
1588
+ "step": 2630
1589
+ },
1590
+ {
1591
+ "grad_norm": 0.24847416579723358,
1592
+ "learning_rate": 6.389661961682173e-05,
1593
+ "loss": 0.0315,
1594
+ "step": 2640
1595
+ },
1596
+ {
1597
+ "grad_norm": 0.31609058380126953,
1598
+ "learning_rate": 6.363168919272846e-05,
1599
+ "loss": 0.039,
1600
+ "step": 2650
1601
+ },
1602
+ {
1603
+ "grad_norm": 0.2490130513906479,
1604
+ "learning_rate": 6.336634467494768e-05,
1605
+ "loss": 0.0414,
1606
+ "step": 2660
1607
+ },
1608
+ {
1609
+ "grad_norm": 0.26495233178138733,
1610
+ "learning_rate": 6.310059412392505e-05,
1611
+ "loss": 0.0372,
1612
+ "step": 2670
1613
+ },
1614
+ {
1615
+ "grad_norm": 0.30787259340286255,
1616
+ "learning_rate": 6.283444561244042e-05,
1617
+ "loss": 0.0396,
1618
+ "step": 2680
1619
+ },
1620
+ {
1621
+ "grad_norm": 0.29595232009887695,
1622
+ "learning_rate": 6.256790722536251e-05,
1623
+ "loss": 0.0364,
1624
+ "step": 2690
1625
+ },
1626
+ {
1627
+ "grad_norm": 0.30548182129859924,
1628
+ "learning_rate": 6.230098705940354e-05,
1629
+ "loss": 0.0384,
1630
+ "step": 2700
1631
+ },
1632
+ {
1633
+ "grad_norm": 0.1984514445066452,
1634
+ "learning_rate": 6.203369322287306e-05,
1635
+ "loss": 0.0342,
1636
+ "step": 2710
1637
+ },
1638
+ {
1639
+ "grad_norm": 0.2916681468486786,
1640
+ "learning_rate": 6.17660338354317e-05,
1641
+ "loss": 0.0315,
1642
+ "step": 2720
1643
+ },
1644
+ {
1645
+ "grad_norm": 0.26730629801750183,
1646
+ "learning_rate": 6.149801702784456e-05,
1647
+ "loss": 0.0327,
1648
+ "step": 2730
1649
+ },
1650
+ {
1651
+ "grad_norm": 0.20792292058467865,
1652
+ "learning_rate": 6.122965094173424e-05,
1653
+ "loss": 0.0351,
1654
+ "step": 2740
1655
+ },
1656
+ {
1657
+ "grad_norm": 0.2516806721687317,
1658
+ "learning_rate": 6.0960943729333374e-05,
1659
+ "loss": 0.0307,
1660
+ "step": 2750
1661
+ },
1662
+ {
1663
+ "grad_norm": 0.32123464345932007,
1664
+ "learning_rate": 6.069190355323717e-05,
1665
+ "loss": 0.0324,
1666
+ "step": 2760
1667
+ },
1668
+ {
1669
+ "grad_norm": 0.282929003238678,
1670
+ "learning_rate": 6.042253858615532e-05,
1671
+ "loss": 0.0321,
1672
+ "step": 2770
1673
+ },
1674
+ {
1675
+ "grad_norm": 0.288335919380188,
1676
+ "learning_rate": 6.015285701066382e-05,
1677
+ "loss": 0.0318,
1678
+ "step": 2780
1679
+ },
1680
+ {
1681
+ "grad_norm": 0.2876138389110565,
1682
+ "learning_rate": 5.988286701895631e-05,
1683
+ "loss": 0.0299,
1684
+ "step": 2790
1685
+ },
1686
+ {
1687
+ "grad_norm": 0.2926459014415741,
1688
+ "learning_rate": 5.961257681259535e-05,
1689
+ "loss": 0.0332,
1690
+ "step": 2800
1691
+ },
1692
+ {
1693
+ "grad_norm": 0.242830291390419,
1694
+ "learning_rate": 5.934199460226317e-05,
1695
+ "loss": 0.0324,
1696
+ "step": 2810
1697
+ },
1698
+ {
1699
+ "grad_norm": 0.34635767340660095,
1700
+ "learning_rate": 5.9071128607512285e-05,
1701
+ "loss": 0.0319,
1702
+ "step": 2820
1703
+ },
1704
+ {
1705
+ "grad_norm": 0.2989971339702606,
1706
+ "learning_rate": 5.8799987056515804e-05,
1707
+ "loss": 0.0341,
1708
+ "step": 2830
1709
+ },
1710
+ {
1711
+ "grad_norm": 0.2593464255332947,
1712
+ "learning_rate": 5.8528578185817514e-05,
1713
+ "loss": 0.0378,
1714
+ "step": 2840
1715
+ },
1716
+ {
1717
+ "grad_norm": 0.3017624020576477,
1718
+ "learning_rate": 5.825691024008162e-05,
1719
+ "loss": 0.0337,
1720
+ "step": 2850
1721
+ },
1722
+ {
1723
+ "grad_norm": 0.23172958195209503,
1724
+ "learning_rate": 5.798499147184233e-05,
1725
+ "loss": 0.0334,
1726
+ "step": 2860
1727
+ },
1728
+ {
1729
+ "grad_norm": 0.3011959195137024,
1730
+ "learning_rate": 5.771283014125317e-05,
1731
+ "loss": 0.0343,
1732
+ "step": 2870
1733
+ },
1734
+ {
1735
+ "grad_norm": 0.19752350449562073,
1736
+ "learning_rate": 5.7440434515836064e-05,
1737
+ "loss": 0.0329,
1738
+ "step": 2880
1739
+ },
1740
+ {
1741
+ "grad_norm": 0.20669515430927277,
1742
+ "learning_rate": 5.7167812870230094e-05,
1743
+ "loss": 0.0325,
1744
+ "step": 2890
1745
+ },
1746
+ {
1747
+ "grad_norm": 0.329367995262146,
1748
+ "learning_rate": 5.689497348594035e-05,
1749
+ "loss": 0.0353,
1750
+ "step": 2900
1751
+ },
1752
+ {
1753
+ "grad_norm": 0.2281043380498886,
1754
+ "learning_rate": 5.662192465108613e-05,
1755
+ "loss": 0.0458,
1756
+ "step": 2910
1757
+ },
1758
+ {
1759
+ "grad_norm": 0.2360934019088745,
1760
+ "learning_rate": 5.634867466014932e-05,
1761
+ "loss": 0.0302,
1762
+ "step": 2920
1763
+ },
1764
+ {
1765
+ "grad_norm": 0.2888210117816925,
1766
+ "learning_rate": 5.607523181372234e-05,
1767
+ "loss": 0.0333,
1768
+ "step": 2930
1769
+ },
1770
+ {
1771
+ "grad_norm": 0.19709353148937225,
1772
+ "learning_rate": 5.5801604418256117e-05,
1773
+ "loss": 0.0326,
1774
+ "step": 2940
1775
+ },
1776
+ {
1777
+ "grad_norm": 0.3951866328716278,
1778
+ "learning_rate": 5.552780078580756e-05,
1779
+ "loss": 0.0331,
1780
+ "step": 2950
1781
+ },
1782
+ {
1783
+ "grad_norm": 0.3455570936203003,
1784
+ "learning_rate": 5.525382923378728e-05,
1785
+ "loss": 0.0358,
1786
+ "step": 2960
1787
+ },
1788
+ {
1789
+ "grad_norm": 0.40430018305778503,
1790
+ "learning_rate": 5.49796980847068e-05,
1791
+ "loss": 0.0355,
1792
+ "step": 2970
1793
+ },
1794
+ {
1795
+ "grad_norm": 0.25933536887168884,
1796
+ "learning_rate": 5.470541566592573e-05,
1797
+ "loss": 0.0303,
1798
+ "step": 2980
1799
+ },
1800
+ {
1801
+ "grad_norm": 0.22030013799667358,
1802
+ "learning_rate": 5.443099030939887e-05,
1803
+ "loss": 0.0296,
1804
+ "step": 2990
1805
+ },
1806
+ {
1807
+ "grad_norm": 0.27321019768714905,
1808
+ "learning_rate": 5.415643035142309e-05,
1809
+ "loss": 0.037,
1810
+ "step": 3000
1811
+ },
1812
+ {
1813
+ "grad_norm": 0.4012751281261444,
1814
+ "learning_rate": 5.3881744132384104e-05,
1815
+ "loss": 0.0305,
1816
+ "step": 3010
1817
+ },
1818
+ {
1819
+ "grad_norm": 0.2611137628555298,
1820
+ "learning_rate": 5.360693999650303e-05,
1821
+ "loss": 0.0352,
1822
+ "step": 3020
1823
+ },
1824
+ {
1825
+ "grad_norm": 0.24199552834033966,
1826
+ "learning_rate": 5.3332026291583016e-05,
1827
+ "loss": 0.0278,
1828
+ "step": 3030
1829
+ },
1830
+ {
1831
+ "grad_norm": 0.16219401359558105,
1832
+ "learning_rate": 5.305701136875566e-05,
1833
+ "loss": 0.026,
1834
+ "step": 3040
1835
+ },
1836
+ {
1837
+ "grad_norm": 0.25862357020378113,
1838
+ "learning_rate": 5.278190358222721e-05,
1839
+ "loss": 0.033,
1840
+ "step": 3050
1841
+ },
1842
+ {
1843
+ "grad_norm": 0.1735977679491043,
1844
+ "learning_rate": 5.25067112890249e-05,
1845
+ "loss": 0.0328,
1846
+ "step": 3060
1847
+ },
1848
+ {
1849
+ "grad_norm": 0.19945716857910156,
1850
+ "learning_rate": 5.2231442848743064e-05,
1851
+ "loss": 0.0341,
1852
+ "step": 3070
1853
+ },
1854
+ {
1855
+ "grad_norm": 0.19215127825737,
1856
+ "learning_rate": 5.1956106623289145e-05,
1857
+ "loss": 0.0289,
1858
+ "step": 3080
1859
+ },
1860
+ {
1861
+ "grad_norm": 0.24303321540355682,
1862
+ "learning_rate": 5.168071097662972e-05,
1863
+ "loss": 0.0344,
1864
+ "step": 3090
1865
+ },
1866
+ {
1867
+ "grad_norm": 0.20462214946746826,
1868
+ "learning_rate": 5.1405264274536445e-05,
1869
+ "loss": 0.0305,
1870
+ "step": 3100
1871
+ },
1872
+ {
1873
+ "grad_norm": 0.19431976974010468,
1874
+ "learning_rate": 5.112977488433188e-05,
1875
+ "loss": 0.0389,
1876
+ "step": 3110
1877
+ },
1878
+ {
1879
+ "grad_norm": 0.1659228801727295,
1880
+ "learning_rate": 5.085425117463533e-05,
1881
+ "loss": 0.032,
1882
+ "step": 3120
1883
+ },
1884
+ {
1885
+ "grad_norm": 0.3553987145423889,
1886
+ "learning_rate": 5.057870151510864e-05,
1887
+ "loss": 0.0354,
1888
+ "step": 3130
1889
+ },
1890
+ {
1891
+ "grad_norm": 0.17798061668872833,
1892
+ "learning_rate": 5.030313427620197e-05,
1893
+ "loss": 0.0283,
1894
+ "step": 3140
1895
+ },
1896
+ {
1897
+ "grad_norm": 0.2874560058116913,
1898
+ "learning_rate": 5.0027557828899426e-05,
1899
+ "loss": 0.0302,
1900
+ "step": 3150
1901
+ },
1902
+ {
1903
+ "grad_norm": 0.22314925491809845,
1904
+ "learning_rate": 4.975198054446492e-05,
1905
+ "loss": 0.0292,
1906
+ "step": 3160
1907
+ },
1908
+ {
1909
+ "grad_norm": 0.21092987060546875,
1910
+ "learning_rate": 4.947641079418773e-05,
1911
+ "loss": 0.0308,
1912
+ "step": 3170
1913
+ },
1914
+ {
1915
+ "grad_norm": 0.2341562956571579,
1916
+ "learning_rate": 4.920085694912828e-05,
1917
+ "loss": 0.0265,
1918
+ "step": 3180
1919
+ },
1920
+ {
1921
+ "grad_norm": 0.2401692271232605,
1922
+ "learning_rate": 4.892532737986387e-05,
1923
+ "loss": 0.0297,
1924
+ "step": 3190
1925
+ },
1926
+ {
1927
+ "grad_norm": 0.3158665597438812,
1928
+ "learning_rate": 4.864983045623434e-05,
1929
+ "loss": 0.0333,
1930
+ "step": 3200
1931
+ },
1932
+ {
1933
+ "grad_norm": 0.3427079916000366,
1934
+ "learning_rate": 4.837437454708784e-05,
1935
+ "loss": 0.0289,
1936
+ "step": 3210
1937
+ },
1938
+ {
1939
+ "grad_norm": 0.2586159408092499,
1940
+ "learning_rate": 4.809896802002662e-05,
1941
+ "loss": 0.0266,
1942
+ "step": 3220
1943
+ },
1944
+ {
1945
+ "grad_norm": 0.23725765943527222,
1946
+ "learning_rate": 4.7823619241152854e-05,
1947
+ "loss": 0.0291,
1948
+ "step": 3230
1949
+ },
1950
+ {
1951
+ "grad_norm": 0.3046949803829193,
1952
+ "learning_rate": 4.754833657481445e-05,
1953
+ "loss": 0.027,
1954
+ "step": 3240
1955
+ },
1956
+ {
1957
+ "grad_norm": 0.24730272591114044,
1958
+ "learning_rate": 4.7273128383351015e-05,
1959
+ "loss": 0.027,
1960
+ "step": 3250
1961
+ },
1962
+ {
1963
+ "grad_norm": 0.2536727786064148,
1964
+ "learning_rate": 4.699800302683981e-05,
1965
+ "loss": 0.0347,
1966
+ "step": 3260
1967
+ },
1968
+ {
1969
+ "grad_norm": 0.25115495920181274,
1970
+ "learning_rate": 4.6722968862841806e-05,
1971
+ "loss": 0.0294,
1972
+ "step": 3270
1973
+ },
1974
+ {
1975
+ "grad_norm": 0.234017014503479,
1976
+ "learning_rate": 4.6448034246147754e-05,
1977
+ "loss": 0.0278,
1978
+ "step": 3280
1979
+ },
1980
+ {
1981
+ "grad_norm": 0.19399166107177734,
1982
+ "learning_rate": 4.6173207528524476e-05,
1983
+ "loss": 0.029,
1984
+ "step": 3290
1985
+ },
1986
+ {
1987
+ "grad_norm": 0.28640416264533997,
1988
+ "learning_rate": 4.58984970584611e-05,
1989
+ "loss": 0.0259,
1990
+ "step": 3300
1991
+ },
1992
+ {
1993
+ "grad_norm": 0.26729804277420044,
1994
+ "learning_rate": 4.562391118091544e-05,
1995
+ "loss": 0.0275,
1996
+ "step": 3310
1997
+ },
1998
+ {
1999
+ "grad_norm": 0.20602551102638245,
2000
+ "learning_rate": 4.534945823706056e-05,
2001
+ "loss": 0.029,
2002
+ "step": 3320
2003
+ },
2004
+ {
2005
+ "grad_norm": 0.3043207824230194,
2006
+ "learning_rate": 4.507514656403137e-05,
2007
+ "loss": 0.0277,
2008
+ "step": 3330
2009
+ },
2010
+ {
2011
+ "grad_norm": 0.18039357662200928,
2012
+ "learning_rate": 4.480098449467132e-05,
2013
+ "loss": 0.026,
2014
+ "step": 3340
2015
+ },
2016
+ {
2017
+ "grad_norm": 0.21921129524707794,
2018
+ "learning_rate": 4.452698035727929e-05,
2019
+ "loss": 0.0306,
2020
+ "step": 3350
2021
+ },
2022
+ {
2023
+ "grad_norm": 0.24986359477043152,
2024
+ "learning_rate": 4.425314247535668e-05,
2025
+ "loss": 0.0267,
2026
+ "step": 3360
2027
+ },
2028
+ {
2029
+ "grad_norm": 0.33457106351852417,
2030
+ "learning_rate": 4.3979479167354477e-05,
2031
+ "loss": 0.0373,
2032
+ "step": 3370
2033
+ },
2034
+ {
2035
+ "grad_norm": 0.30760475993156433,
2036
+ "learning_rate": 4.370599874642055e-05,
2037
+ "loss": 0.0323,
2038
+ "step": 3380
2039
+ },
2040
+ {
2041
+ "grad_norm": 0.26259493827819824,
2042
+ "learning_rate": 4.3432709520147205e-05,
2043
+ "loss": 0.0302,
2044
+ "step": 3390
2045
+ },
2046
+ {
2047
+ "grad_norm": 0.2644908130168915,
2048
+ "learning_rate": 4.315961979031875e-05,
2049
+ "loss": 0.0323,
2050
+ "step": 3400
2051
+ },
2052
+ {
2053
+ "grad_norm": 0.24664807319641113,
2054
+ "learning_rate": 4.2886737852659325e-05,
2055
+ "loss": 0.0282,
2056
+ "step": 3410
2057
+ },
2058
+ {
2059
+ "grad_norm": 0.2367314249277115,
2060
+ "learning_rate": 4.261407199658093e-05,
2061
+ "loss": 0.0271,
2062
+ "step": 3420
2063
+ },
2064
+ {
2065
+ "grad_norm": 0.32822489738464355,
2066
+ "learning_rate": 4.234163050493158e-05,
2067
+ "loss": 0.0253,
2068
+ "step": 3430
2069
+ },
2070
+ {
2071
+ "grad_norm": 0.2989010214805603,
2072
+ "learning_rate": 4.2069421653743706e-05,
2073
+ "loss": 0.0285,
2074
+ "step": 3440
2075
+ },
2076
+ {
2077
+ "grad_norm": 0.2728559076786041,
2078
+ "learning_rate": 4.179745371198276e-05,
2079
+ "loss": 0.0241,
2080
+ "step": 3450
2081
+ },
2082
+ {
2083
+ "grad_norm": 0.20925480127334595,
2084
+ "learning_rate": 4.1525734941296026e-05,
2085
+ "loss": 0.0291,
2086
+ "step": 3460
2087
+ },
2088
+ {
2089
+ "grad_norm": 0.2235332429409027,
2090
+ "learning_rate": 4.125427359576162e-05,
2091
+ "loss": 0.0244,
2092
+ "step": 3470
2093
+ },
2094
+ {
2095
+ "grad_norm": 0.23195625841617584,
2096
+ "learning_rate": 4.0983077921637815e-05,
2097
+ "loss": 0.0269,
2098
+ "step": 3480
2099
+ },
2100
+ {
2101
+ "grad_norm": 0.282319575548172,
2102
+ "learning_rate": 4.07121561571125e-05,
2103
+ "loss": 0.0263,
2104
+ "step": 3490
2105
+ },
2106
+ {
2107
+ "grad_norm": 0.29623618721961975,
2108
+ "learning_rate": 4.044151653205292e-05,
2109
+ "loss": 0.0312,
2110
+ "step": 3500
2111
+ },
2112
+ {
2113
+ "grad_norm": 0.22592341899871826,
2114
+ "learning_rate": 4.0171167267755696e-05,
2115
+ "loss": 0.0294,
2116
+ "step": 3510
2117
+ },
2118
+ {
2119
+ "grad_norm": 0.21023550629615784,
2120
+ "learning_rate": 3.9901116576697083e-05,
2121
+ "loss": 0.0246,
2122
+ "step": 3520
2123
+ },
2124
+ {
2125
+ "grad_norm": 0.21181227266788483,
2126
+ "learning_rate": 3.963137266228349e-05,
2127
+ "loss": 0.0249,
2128
+ "step": 3530
2129
+ },
2130
+ {
2131
+ "grad_norm": 0.30305394530296326,
2132
+ "learning_rate": 3.93619437186023e-05,
2133
+ "loss": 0.0313,
2134
+ "step": 3540
2135
+ },
2136
+ {
2137
+ "grad_norm": 0.21164947748184204,
2138
+ "learning_rate": 3.9092837930172884e-05,
2139
+ "loss": 0.0262,
2140
+ "step": 3550
2141
+ },
2142
+ {
2143
+ "grad_norm": 0.29877039790153503,
2144
+ "learning_rate": 3.8824063471698105e-05,
2145
+ "loss": 0.028,
2146
+ "step": 3560
2147
+ },
2148
+ {
2149
+ "grad_norm": 0.31443488597869873,
2150
+ "learning_rate": 3.855562850781589e-05,
2151
+ "loss": 0.0292,
2152
+ "step": 3570
2153
+ },
2154
+ {
2155
+ "grad_norm": 0.26347023248672485,
2156
+ "learning_rate": 3.828754119285123e-05,
2157
+ "loss": 0.0273,
2158
+ "step": 3580
2159
+ },
2160
+ {
2161
+ "grad_norm": 0.23366351425647736,
2162
+ "learning_rate": 3.801980967056851e-05,
2163
+ "loss": 0.0253,
2164
+ "step": 3590
2165
+ },
2166
+ {
2167
+ "grad_norm": 0.2442861944437027,
2168
+ "learning_rate": 3.77524420739241e-05,
2169
+ "loss": 0.0274,
2170
+ "step": 3600
2171
+ },
2172
+ {
2173
+ "grad_norm": 0.23381778597831726,
2174
+ "learning_rate": 3.748544652481927e-05,
2175
+ "loss": 0.022,
2176
+ "step": 3610
2177
+ },
2178
+ {
2179
+ "grad_norm": 0.17044523358345032,
2180
+ "learning_rate": 3.721883113385353e-05,
2181
+ "loss": 0.0275,
2182
+ "step": 3620
2183
+ },
2184
+ {
2185
+ "grad_norm": 0.1821439266204834,
2186
+ "learning_rate": 3.695260400007819e-05,
2187
+ "loss": 0.0274,
2188
+ "step": 3630
2189
+ },
2190
+ {
2191
+ "grad_norm": 0.272661954164505,
2192
+ "learning_rate": 3.6686773210750385e-05,
2193
+ "loss": 0.0279,
2194
+ "step": 3640
2195
+ },
2196
+ {
2197
+ "grad_norm": 0.25035589933395386,
2198
+ "learning_rate": 3.642134684108737e-05,
2199
+ "loss": 0.0306,
2200
+ "step": 3650
2201
+ },
2202
+ {
2203
+ "grad_norm": 0.27742305397987366,
2204
+ "learning_rate": 3.615633295402123e-05,
2205
+ "loss": 0.0217,
2206
+ "step": 3660
2207
+ },
2208
+ {
2209
+ "grad_norm": 0.18357422947883606,
2210
+ "learning_rate": 3.5891739599953945e-05,
2211
+ "loss": 0.0264,
2212
+ "step": 3670
2213
+ },
2214
+ {
2215
+ "grad_norm": 0.2029154896736145,
2216
+ "learning_rate": 3.5627574816512846e-05,
2217
+ "loss": 0.0273,
2218
+ "step": 3680
2219
+ },
2220
+ {
2221
+ "grad_norm": 0.25645092129707336,
2222
+ "learning_rate": 3.536384662830648e-05,
2223
+ "loss": 0.0293,
2224
+ "step": 3690
2225
+ },
2226
+ {
2227
+ "grad_norm": 0.2826383709907532,
2228
+ "learning_rate": 3.5100563046680764e-05,
2229
+ "loss": 0.0249,
2230
+ "step": 3700
2231
+ },
2232
+ {
2233
+ "grad_norm": 0.22233398258686066,
2234
+ "learning_rate": 3.483773206947572e-05,
2235
+ "loss": 0.028,
2236
+ "step": 3710
2237
+ },
2238
+ {
2239
+ "grad_norm": 0.21906347572803497,
2240
+ "learning_rate": 3.457536168078247e-05,
2241
+ "loss": 0.0268,
2242
+ "step": 3720
2243
+ },
2244
+ {
2245
+ "grad_norm": 0.22737747430801392,
2246
+ "learning_rate": 3.431345985070067e-05,
2247
+ "loss": 0.0231,
2248
+ "step": 3730
2249
+ },
2250
+ {
2251
+ "grad_norm": 0.24586769938468933,
2252
+ "learning_rate": 3.40520345350965e-05,
2253
+ "loss": 0.0287,
2254
+ "step": 3740
2255
+ },
2256
+ {
2257
+ "grad_norm": 0.1879250854253769,
2258
+ "learning_rate": 3.379109367536089e-05,
2259
+ "loss": 0.0271,
2260
+ "step": 3750
2261
+ },
2262
+ {
2263
+ "grad_norm": 0.15110410749912262,
2264
+ "learning_rate": 3.3530645198168295e-05,
2265
+ "loss": 0.0252,
2266
+ "step": 3760
2267
+ },
2268
+ {
2269
+ "grad_norm": 0.24863441288471222,
2270
+ "learning_rate": 3.327069701523595e-05,
2271
+ "loss": 0.0296,
2272
+ "step": 3770
2273
+ },
2274
+ {
2275
+ "grad_norm": 0.25430187582969666,
2276
+ "learning_rate": 3.301125702308353e-05,
2277
+ "loss": 0.0255,
2278
+ "step": 3780
2279
+ },
2280
+ {
2281
+ "grad_norm": 0.21225595474243164,
2282
+ "learning_rate": 3.275233310279321e-05,
2283
+ "loss": 0.0238,
2284
+ "step": 3790
2285
+ },
2286
+ {
2287
+ "grad_norm": 0.1806585192680359,
2288
+ "learning_rate": 3.249393311977037e-05,
2289
+ "loss": 0.0258,
2290
+ "step": 3800
2291
+ },
2292
+ {
2293
+ "grad_norm": 0.25602349638938904,
2294
+ "learning_rate": 3.223606492350451e-05,
2295
+ "loss": 0.0232,
2296
+ "step": 3810
2297
+ },
2298
+ {
2299
+ "grad_norm": 0.19176100194454193,
2300
+ "learning_rate": 3.197873634733096e-05,
2301
+ "loss": 0.026,
2302
+ "step": 3820
2303
+ },
2304
+ {
2305
+ "grad_norm": 0.22703717648983002,
2306
+ "learning_rate": 3.172195520819285e-05,
2307
+ "loss": 0.0282,
2308
+ "step": 3830
2309
+ },
2310
+ {
2311
+ "grad_norm": 0.15395799279212952,
2312
+ "learning_rate": 3.146572930640362e-05,
2313
+ "loss": 0.0277,
2314
+ "step": 3840
2315
+ },
2316
+ {
2317
+ "grad_norm": 0.1866094321012497,
2318
+ "learning_rate": 3.121006642541014e-05,
2319
+ "loss": 0.023,
2320
+ "step": 3850
2321
+ },
2322
+ {
2323
+ "grad_norm": 0.20458878576755524,
2324
+ "learning_rate": 3.095497433155626e-05,
2325
+ "loss": 0.0279,
2326
+ "step": 3860
2327
+ },
2328
+ {
2329
+ "grad_norm": 0.32518133521080017,
2330
+ "learning_rate": 3.070046077384682e-05,
2331
+ "loss": 0.0269,
2332
+ "step": 3870
2333
+ },
2334
+ {
2335
+ "grad_norm": 0.18735451996326447,
2336
+ "learning_rate": 3.0446533483712304e-05,
2337
+ "loss": 0.0236,
2338
+ "step": 3880
2339
+ },
2340
+ {
2341
+ "grad_norm": 0.1771942526102066,
2342
+ "learning_rate": 3.0193200174774038e-05,
2343
+ "loss": 0.0222,
2344
+ "step": 3890
2345
+ },
2346
+ {
2347
+ "grad_norm": 0.1862131953239441,
2348
+ "learning_rate": 2.994046854260974e-05,
2349
+ "loss": 0.0293,
2350
+ "step": 3900
2351
+ },
2352
+ {
2353
+ "grad_norm": 0.15679983794689178,
2354
+ "learning_rate": 2.9688346264519866e-05,
2355
+ "loss": 0.0235,
2356
+ "step": 3910
2357
+ },
2358
+ {
2359
+ "grad_norm": 0.16667503118515015,
2360
+ "learning_rate": 2.943684099929436e-05,
2361
+ "loss": 0.0233,
2362
+ "step": 3920
2363
+ },
2364
+ {
2365
+ "grad_norm": 0.22355644404888153,
2366
+ "learning_rate": 2.918596038697995e-05,
2367
+ "loss": 0.0249,
2368
+ "step": 3930
2369
+ },
2370
+ {
2371
+ "grad_norm": 0.22866907715797424,
2372
+ "learning_rate": 2.8935712048648112e-05,
2373
+ "loss": 0.0285,
2374
+ "step": 3940
2375
+ },
2376
+ {
2377
+ "grad_norm": 0.18918220698833466,
2378
+ "learning_rate": 2.8686103586163626e-05,
2379
+ "loss": 0.0261,
2380
+ "step": 3950
2381
+ },
2382
+ {
2383
+ "grad_norm": 0.21910104155540466,
2384
+ "learning_rate": 2.843714258195346e-05,
2385
+ "loss": 0.0227,
2386
+ "step": 3960
2387
+ },
2388
+ {
2389
+ "grad_norm": 0.2601633071899414,
2390
+ "learning_rate": 2.8188836598776662e-05,
2391
+ "loss": 0.0277,
2392
+ "step": 3970
2393
+ },
2394
+ {
2395
+ "grad_norm": 0.1436263620853424,
2396
+ "learning_rate": 2.7941193179494484e-05,
2397
+ "loss": 0.0226,
2398
+ "step": 3980
2399
+ },
2400
+ {
2401
+ "grad_norm": 0.15300126373767853,
2402
+ "learning_rate": 2.7694219846841262e-05,
2403
+ "loss": 0.0242,
2404
+ "step": 3990
2405
+ },
2406
+ {
2407
+ "grad_norm": 0.1818075329065323,
2408
+ "learning_rate": 2.7447924103195976e-05,
2409
+ "loss": 0.0271,
2410
+ "step": 4000
2411
+ },
2412
+ {
2413
+ "grad_norm": 0.2131088674068451,
2414
+ "learning_rate": 2.7202313430354253e-05,
2415
+ "loss": 0.0217,
2416
+ "step": 4010
2417
+ },
2418
+ {
2419
+ "grad_norm": 0.2225767970085144,
2420
+ "learning_rate": 2.695739528930111e-05,
2421
+ "loss": 0.0231,
2422
+ "step": 4020
2423
+ },
2424
+ {
2425
+ "grad_norm": 0.2219320386648178,
2426
+ "learning_rate": 2.67131771199844e-05,
2427
+ "loss": 0.0255,
2428
+ "step": 4030
2429
+ },
2430
+ {
2431
+ "grad_norm": 0.19305405020713806,
2432
+ "learning_rate": 2.6469666341088677e-05,
2433
+ "loss": 0.0241,
2434
+ "step": 4040
2435
+ },
2436
+ {
2437
+ "grad_norm": 0.1855151206254959,
2438
+ "learning_rate": 2.6226870349809885e-05,
2439
+ "loss": 0.0213,
2440
+ "step": 4050
2441
+ },
2442
+ {
2443
+ "grad_norm": 0.21167217195034027,
2444
+ "learning_rate": 2.5984796521630737e-05,
2445
+ "loss": 0.0257,
2446
+ "step": 4060
2447
+ },
2448
+ {
2449
+ "grad_norm": 0.18498139083385468,
2450
+ "learning_rate": 2.574345221009653e-05,
2451
+ "loss": 0.0244,
2452
+ "step": 4070
2453
+ },
2454
+ {
2455
+ "grad_norm": 0.2016761749982834,
2456
+ "learning_rate": 2.5502844746591804e-05,
2457
+ "loss": 0.0211,
2458
+ "step": 4080
2459
+ },
2460
+ {
2461
+ "grad_norm": 0.225340336561203,
2462
+ "learning_rate": 2.526298144011775e-05,
2463
+ "loss": 0.0234,
2464
+ "step": 4090
2465
+ },
2466
+ {
2467
+ "grad_norm": 0.23794017732143402,
2468
+ "learning_rate": 2.5023869577070013e-05,
2469
+ "loss": 0.0261,
2470
+ "step": 4100
2471
+ },
2472
+ {
2473
+ "grad_norm": 0.1910785734653473,
2474
+ "learning_rate": 2.478551642101743e-05,
2475
+ "loss": 0.0233,
2476
+ "step": 4110
2477
+ },
2478
+ {
2479
+ "grad_norm": 0.211565762758255,
2480
+ "learning_rate": 2.4547929212481435e-05,
2481
+ "loss": 0.0193,
2482
+ "step": 4120
2483
+ },
2484
+ {
2485
+ "grad_norm": 0.18004284799098969,
2486
+ "learning_rate": 2.4311115168716013e-05,
2487
+ "loss": 0.0217,
2488
+ "step": 4130
2489
+ },
2490
+ {
2491
+ "grad_norm": 0.2306419312953949,
2492
+ "learning_rate": 2.4075081483488494e-05,
2493
+ "loss": 0.0269,
2494
+ "step": 4140
2495
+ },
2496
+ {
2497
+ "grad_norm": 0.20246738195419312,
2498
+ "learning_rate": 2.3839835326861104e-05,
2499
+ "loss": 0.0215,
2500
+ "step": 4150
2501
+ },
2502
+ {
2503
+ "grad_norm": 0.18215186893939972,
2504
+ "learning_rate": 2.3605383844972966e-05,
2505
+ "loss": 0.0199,
2506
+ "step": 4160
2507
+ },
2508
+ {
2509
+ "grad_norm": 0.2206883728504181,
2510
+ "learning_rate": 2.3371734159823284e-05,
2511
+ "loss": 0.0259,
2512
+ "step": 4170
2513
+ },
2514
+ {
2515
+ "grad_norm": 0.24595214426517487,
2516
+ "learning_rate": 2.3138893369054766e-05,
2517
+ "loss": 0.0268,
2518
+ "step": 4180
2519
+ },
2520
+ {
2521
+ "grad_norm": 0.2135511040687561,
2522
+ "learning_rate": 2.2906868545738102e-05,
2523
+ "loss": 0.0253,
2524
+ "step": 4190
2525
+ },
2526
+ {
2527
+ "grad_norm": 0.18752355873584747,
2528
+ "learning_rate": 2.2675666738157186e-05,
2529
+ "loss": 0.0186,
2530
+ "step": 4200
2531
+ },
2532
+ {
2533
+ "grad_norm": 0.17715811729431152,
2534
+ "learning_rate": 2.2445294969594844e-05,
2535
+ "loss": 0.0171,
2536
+ "step": 4210
2537
+ },
2538
+ {
2539
+ "grad_norm": 0.2948547303676605,
2540
+ "learning_rate": 2.22157602381196e-05,
2541
+ "loss": 0.02,
2542
+ "step": 4220
2543
+ },
2544
+ {
2545
+ "grad_norm": 0.12416701763868332,
2546
+ "learning_rate": 2.1987069516373098e-05,
2547
+ "loss": 0.0202,
2548
+ "step": 4230
2549
+ },
2550
+ {
2551
+ "grad_norm": 0.1834097057580948,
2552
+ "learning_rate": 2.1759229751358217e-05,
2553
+ "loss": 0.0252,
2554
+ "step": 4240
2555
+ },
2556
+ {
2557
+ "grad_norm": 0.16251210868358612,
2558
+ "learning_rate": 2.1532247864228084e-05,
2559
+ "loss": 0.0226,
2560
+ "step": 4250
2561
+ },
2562
+ {
2563
+ "grad_norm": 0.11320555210113525,
2564
+ "learning_rate": 2.1306130750075865e-05,
2565
+ "loss": 0.0264,
2566
+ "step": 4260
2567
+ },
2568
+ {
2569
+ "grad_norm": 0.16259628534317017,
2570
+ "learning_rate": 2.1080885277725236e-05,
2571
+ "loss": 0.0203,
2572
+ "step": 4270
2573
+ },
2574
+ {
2575
+ "grad_norm": 0.16590796411037445,
2576
+ "learning_rate": 2.085651828952175e-05,
2577
+ "loss": 0.0218,
2578
+ "step": 4280
2579
+ },
2580
+ {
2581
+ "grad_norm": 0.19988055527210236,
2582
+ "learning_rate": 2.063303660112506e-05,
2583
+ "loss": 0.0237,
2584
+ "step": 4290
2585
+ },
2586
+ {
2587
+ "grad_norm": 0.2420492172241211,
2588
+ "learning_rate": 2.0410447001301753e-05,
2589
+ "loss": 0.0227,
2590
+ "step": 4300
2591
+ },
2592
+ {
2593
+ "grad_norm": 0.2309999167919159,
2594
+ "learning_rate": 2.0188756251719203e-05,
2595
+ "loss": 0.0194,
2596
+ "step": 4310
2597
+ },
2598
+ {
2599
+ "grad_norm": 0.18164952099323273,
2600
+ "learning_rate": 1.9967971086740195e-05,
2601
+ "loss": 0.0209,
2602
+ "step": 4320
2603
+ },
2604
+ {
2605
+ "grad_norm": 0.2178877294063568,
2606
+ "learning_rate": 1.974809821321827e-05,
2607
+ "loss": 0.0198,
2608
+ "step": 4330
2609
+ },
2610
+ {
2611
+ "grad_norm": 0.21368789672851562,
2612
+ "learning_rate": 1.9529144310294023e-05,
2613
+ "loss": 0.0198,
2614
+ "step": 4340
2615
+ },
2616
+ {
2617
+ "grad_norm": 0.22808904945850372,
2618
+ "learning_rate": 1.9311116029192278e-05,
2619
+ "loss": 0.0219,
2620
+ "step": 4350
2621
+ },
2622
+ {
2623
+ "grad_norm": 0.2036302238702774,
2624
+ "learning_rate": 1.909401999301993e-05,
2625
+ "loss": 0.0178,
2626
+ "step": 4360
2627
+ },
2628
+ {
2629
+ "grad_norm": 0.2856689989566803,
2630
+ "learning_rate": 1.887786279656482e-05,
2631
+ "loss": 0.0213,
2632
+ "step": 4370
2633
+ },
2634
+ {
2635
+ "grad_norm": 0.18134640157222748,
2636
+ "learning_rate": 1.8662651006095387e-05,
2637
+ "loss": 0.0201,
2638
+ "step": 4380
2639
+ },
2640
+ {
2641
+ "grad_norm": 0.278046190738678,
2642
+ "learning_rate": 1.8448391159161204e-05,
2643
+ "loss": 0.0244,
2644
+ "step": 4390
2645
+ },
2646
+ {
2647
+ "grad_norm": 0.26307907700538635,
2648
+ "learning_rate": 1.8235089764394408e-05,
2649
+ "loss": 0.0274,
2650
+ "step": 4400
2651
+ },
2652
+ {
2653
+ "grad_norm": 0.22492247819900513,
2654
+ "learning_rate": 1.8022753301311935e-05,
2655
+ "loss": 0.0212,
2656
+ "step": 4410
2657
+ },
2658
+ {
2659
+ "grad_norm": 0.19130860269069672,
2660
+ "learning_rate": 1.7811388220118707e-05,
2661
+ "loss": 0.0237,
2662
+ "step": 4420
2663
+ },
2664
+ {
2665
+ "grad_norm": 0.18481294810771942,
2666
+ "learning_rate": 1.7601000941511757e-05,
2667
+ "loss": 0.0232,
2668
+ "step": 4430
2669
+ },
2670
+ {
2671
+ "grad_norm": 0.1777241826057434,
2672
+ "learning_rate": 1.7391597856485083e-05,
2673
+ "loss": 0.0245,
2674
+ "step": 4440
2675
+ },
2676
+ {
2677
+ "grad_norm": 0.2160668522119522,
2678
+ "learning_rate": 1.7183185326135543e-05,
2679
+ "loss": 0.0171,
2680
+ "step": 4450
2681
+ },
2682
+ {
2683
+ "grad_norm": 0.20380406081676483,
2684
+ "learning_rate": 1.6975769681469705e-05,
2685
+ "loss": 0.0188,
2686
+ "step": 4460
2687
+ },
2688
+ {
2689
+ "grad_norm": 0.20712777972221375,
2690
+ "learning_rate": 1.676935722321139e-05,
2691
+ "loss": 0.0207,
2692
+ "step": 4470
2693
+ },
2694
+ {
2695
+ "grad_norm": 0.20267875492572784,
2696
+ "learning_rate": 1.6563954221610355e-05,
2697
+ "loss": 0.0191,
2698
+ "step": 4480
2699
+ },
2700
+ {
2701
+ "grad_norm": 0.21463575959205627,
2702
+ "learning_rate": 1.6359566916251845e-05,
2703
+ "loss": 0.0208,
2704
+ "step": 4490
2705
+ },
2706
+ {
2707
+ "grad_norm": 0.1477004885673523,
2708
+ "learning_rate": 1.615620151586697e-05,
2709
+ "loss": 0.0215,
2710
+ "step": 4500
2711
+ },
2712
+ {
2713
+ "grad_norm": 0.21711723506450653,
2714
+ "learning_rate": 1.5953864198144135e-05,
2715
+ "loss": 0.0241,
2716
+ "step": 4510
2717
+ },
2718
+ {
2719
+ "grad_norm": 0.1844751238822937,
2720
+ "learning_rate": 1.5752561109541447e-05,
2721
+ "loss": 0.0208,
2722
+ "step": 4520
2723
+ },
2724
+ {
2725
+ "grad_norm": 0.1820620745420456,
2726
+ "learning_rate": 1.5552298365099882e-05,
2727
+ "loss": 0.0218,
2728
+ "step": 4530
2729
+ },
2730
+ {
2731
+ "grad_norm": 0.23407231271266937,
2732
+ "learning_rate": 1.5353082048257596e-05,
2733
+ "loss": 0.0191,
2734
+ "step": 4540
2735
+ },
2736
+ {
2737
+ "grad_norm": 0.21293601393699646,
2738
+ "learning_rate": 1.5154918210665148e-05,
2739
+ "loss": 0.0195,
2740
+ "step": 4550
2741
+ },
2742
+ {
2743
+ "grad_norm": 0.20897363126277924,
2744
+ "learning_rate": 1.4957812872001614e-05,
2745
+ "loss": 0.0186,
2746
+ "step": 4560
2747
+ },
2748
+ {
2749
+ "grad_norm": 0.13340842723846436,
2750
+ "learning_rate": 1.4761772019791748e-05,
2751
+ "loss": 0.0192,
2752
+ "step": 4570
2753
+ },
2754
+ {
2755
+ "grad_norm": 0.24015949666500092,
2756
+ "learning_rate": 1.4566801609224096e-05,
2757
+ "loss": 0.0187,
2758
+ "step": 4580
2759
+ },
2760
+ {
2761
+ "grad_norm": 0.186085045337677,
2762
+ "learning_rate": 1.4372907562970079e-05,
2763
+ "loss": 0.0193,
2764
+ "step": 4590
2765
+ },
2766
+ {
2767
+ "grad_norm": 0.18757310509681702,
2768
+ "learning_rate": 1.4180095771004154e-05,
2769
+ "loss": 0.0184,
2770
+ "step": 4600
2771
+ },
2772
+ {
2773
+ "grad_norm": 0.2534388303756714,
2774
+ "learning_rate": 1.3988372090424773e-05,
2775
+ "loss": 0.0202,
2776
+ "step": 4610
2777
+ },
2778
+ {
2779
+ "grad_norm": 0.16483885049819946,
2780
+ "learning_rate": 1.3797742345276521e-05,
2781
+ "loss": 0.0243,
2782
+ "step": 4620
2783
+ },
2784
+ {
2785
+ "grad_norm": 0.1554243564605713,
2786
+ "learning_rate": 1.3608212326373249e-05,
2787
+ "loss": 0.0183,
2788
+ "step": 4630
2789
+ },
2790
+ {
2791
+ "grad_norm": 0.22151866555213928,
2792
+ "learning_rate": 1.3419787791122062e-05,
2793
+ "loss": 0.0231,
2794
+ "step": 4640
2795
+ },
2796
+ {
2797
+ "grad_norm": 0.21161040663719177,
2798
+ "learning_rate": 1.323247446334847e-05,
2799
+ "loss": 0.0201,
2800
+ "step": 4650
2801
+ },
2802
+ {
2803
+ "grad_norm": 0.186412513256073,
2804
+ "learning_rate": 1.3046278033122577e-05,
2805
+ "loss": 0.025,
2806
+ "step": 4660
2807
+ },
2808
+ {
2809
+ "grad_norm": 0.18747615814208984,
2810
+ "learning_rate": 1.286120415658611e-05,
2811
+ "loss": 0.0203,
2812
+ "step": 4670
2813
+ },
2814
+ {
2815
+ "grad_norm": 0.1292790323495865,
2816
+ "learning_rate": 1.2677258455780683e-05,
2817
+ "loss": 0.0145,
2818
+ "step": 4680
2819
+ },
2820
+ {
2821
+ "grad_norm": 0.17994563281536102,
2822
+ "learning_rate": 1.2494446518477022e-05,
2823
+ "loss": 0.0231,
2824
+ "step": 4690
2825
+ },
2826
+ {
2827
+ "grad_norm": 0.1702570915222168,
2828
+ "learning_rate": 1.2312773898005175e-05,
2829
+ "loss": 0.0223,
2830
+ "step": 4700
2831
+ },
2832
+ {
2833
+ "grad_norm": 0.13994930684566498,
2834
+ "learning_rate": 1.2132246113085822e-05,
2835
+ "loss": 0.0187,
2836
+ "step": 4710
2837
+ },
2838
+ {
2839
+ "grad_norm": 0.14905129373073578,
2840
+ "learning_rate": 1.1952868647662696e-05,
2841
+ "loss": 0.0219,
2842
+ "step": 4720
2843
+ },
2844
+ {
2845
+ "grad_norm": 0.16883762180805206,
2846
+ "learning_rate": 1.1774646950735913e-05,
2847
+ "loss": 0.0189,
2848
+ "step": 4730
2849
+ },
2850
+ {
2851
+ "grad_norm": 0.23552639782428741,
2852
+ "learning_rate": 1.1597586436196473e-05,
2853
+ "loss": 0.0176,
2854
+ "step": 4740
2855
+ },
2856
+ {
2857
+ "grad_norm": 0.24354813992977142,
2858
+ "learning_rate": 1.1421692482661856e-05,
2859
+ "loss": 0.0183,
2860
+ "step": 4750
2861
+ },
2862
+ {
2863
+ "grad_norm": 0.19600911438465118,
2864
+ "learning_rate": 1.124697043331256e-05,
2865
+ "loss": 0.0209,
2866
+ "step": 4760
2867
+ },
2868
+ {
2869
+ "grad_norm": 0.16714124381542206,
2870
+ "learning_rate": 1.107342559572977e-05,
2871
+ "loss": 0.0202,
2872
+ "step": 4770
2873
+ },
2874
+ {
2875
+ "grad_norm": 0.12511172890663147,
2876
+ "learning_rate": 1.090106324173426e-05,
2877
+ "loss": 0.0204,
2878
+ "step": 4780
2879
+ },
2880
+ {
2881
+ "grad_norm": 0.2145424783229828,
2882
+ "learning_rate": 1.0729888607226113e-05,
2883
+ "loss": 0.0203,
2884
+ "step": 4790
2885
+ },
2886
+ {
2887
+ "grad_norm": 0.21264775097370148,
2888
+ "learning_rate": 1.0559906892025745e-05,
2889
+ "loss": 0.0209,
2890
+ "step": 4800
2891
+ },
2892
+ {
2893
+ "grad_norm": 0.18948693573474884,
2894
+ "learning_rate": 1.0391123259715906e-05,
2895
+ "loss": 0.017,
2896
+ "step": 4810
2897
+ },
2898
+ {
2899
+ "grad_norm": 0.16713643074035645,
2900
+ "learning_rate": 1.0223542837484839e-05,
2901
+ "loss": 0.017,
2902
+ "step": 4820
2903
+ },
2904
+ {
2905
+ "grad_norm": 0.21101868152618408,
2906
+ "learning_rate": 1.0057170715970559e-05,
2907
+ "loss": 0.0241,
2908
+ "step": 4830
2909
+ },
2910
+ {
2911
+ "grad_norm": 0.14855772256851196,
2912
+ "learning_rate": 9.892011949106172e-06,
2913
+ "loss": 0.0248,
2914
+ "step": 4840
2915
+ },
2916
+ {
2917
+ "grad_norm": 0.1694764643907547,
2918
+ "learning_rate": 9.728071553966339e-06,
2919
+ "loss": 0.0199,
2920
+ "step": 4850
2921
+ },
2922
+ {
2923
+ "grad_norm": 0.17370817065238953,
2924
+ "learning_rate": 9.56535451061496e-06,
2925
+ "loss": 0.0193,
2926
+ "step": 4860
2927
+ },
2928
+ {
2929
+ "grad_norm": 0.14265552163124084,
2930
+ "learning_rate": 9.403865761953779e-06,
2931
+ "loss": 0.0176,
2932
+ "step": 4870
2933
+ },
2934
+ {
2935
+ "grad_norm": 0.1497550755739212,
2936
+ "learning_rate": 9.243610213572285e-06,
2937
+ "loss": 0.0228,
2938
+ "step": 4880
2939
+ },
2940
+ {
2941
+ "grad_norm": 0.21949000656604767,
2942
+ "learning_rate": 9.084592733598735e-06,
2943
+ "loss": 0.019,
2944
+ "step": 4890
2945
+ },
2946
+ {
2947
+ "grad_norm": 0.13839106261730194,
2948
+ "learning_rate": 8.92681815255219e-06,
2949
+ "loss": 0.0178,
2950
+ "step": 4900
2951
+ },
2952
+ {
2953
+ "grad_norm": 0.14562994241714478,
2954
+ "learning_rate": 8.770291263195819e-06,
2955
+ "loss": 0.0185,
2956
+ "step": 4910
2957
+ },
2958
+ {
2959
+ "grad_norm": 0.19483783841133118,
2960
+ "learning_rate": 8.615016820391342e-06,
2961
+ "loss": 0.0158,
2962
+ "step": 4920
2963
+ },
2964
+ {
2965
+ "grad_norm": 0.16348987817764282,
2966
+ "learning_rate": 8.460999540954517e-06,
2967
+ "loss": 0.0182,
2968
+ "step": 4930
2969
+ },
2970
+ {
2971
+ "grad_norm": 0.18384204804897308,
2972
+ "learning_rate": 8.308244103511909e-06,
2973
+ "loss": 0.0152,
2974
+ "step": 4940
2975
+ },
2976
+ {
2977
+ "grad_norm": 0.16230851411819458,
2978
+ "learning_rate": 8.156755148358764e-06,
2979
+ "loss": 0.0196,
2980
+ "step": 4950
2981
+ },
2982
+ {
2983
+ "grad_norm": 0.19277586042881012,
2984
+ "learning_rate": 8.00653727731801e-06,
2985
+ "loss": 0.0156,
2986
+ "step": 4960
2987
+ },
2988
+ {
2989
+ "grad_norm": 0.21639972925186157,
2990
+ "learning_rate": 7.857595053600513e-06,
2991
+ "loss": 0.024,
2992
+ "step": 4970
2993
+ },
2994
+ {
2995
+ "grad_norm": 0.21124367415905,
2996
+ "learning_rate": 7.709933001666431e-06,
2997
+ "loss": 0.0224,
2998
+ "step": 4980
2999
+ },
3000
+ {
3001
+ "grad_norm": 0.17293235659599304,
3002
+ "learning_rate": 7.56355560708778e-06,
3003
+ "loss": 0.016,
3004
+ "step": 4990
3005
+ },
3006
+ {
3007
+ "grad_norm": 0.1931348741054535,
3008
+ "learning_rate": 7.418467316412158e-06,
3009
+ "loss": 0.0177,
3010
+ "step": 5000
3011
+ },
3012
+ {
3013
+ "grad_norm": 0.14251281321048737,
3014
+ "learning_rate": 7.2746725370277435e-06,
3015
+ "loss": 0.0186,
3016
+ "step": 5010
3017
+ },
3018
+ {
3019
+ "grad_norm": 0.15622611343860626,
3020
+ "learning_rate": 7.132175637029293e-06,
3021
+ "loss": 0.02,
3022
+ "step": 5020
3023
+ },
3024
+ {
3025
+ "grad_norm": 0.18546737730503082,
3026
+ "learning_rate": 6.9909809450855345e-06,
3027
+ "loss": 0.0193,
3028
+ "step": 5030
3029
+ },
3030
+ {
3031
+ "grad_norm": 0.18201150000095367,
3032
+ "learning_rate": 6.851092750307686e-06,
3033
+ "loss": 0.0159,
3034
+ "step": 5040
3035
+ },
3036
+ {
3037
+ "grad_norm": 0.20087608695030212,
3038
+ "learning_rate": 6.712515302119077e-06,
3039
+ "loss": 0.0185,
3040
+ "step": 5050
3041
+ },
3042
+ {
3043
+ "grad_norm": 0.21938590705394745,
3044
+ "learning_rate": 6.575252810126143e-06,
3045
+ "loss": 0.0181,
3046
+ "step": 5060
3047
+ },
3048
+ {
3049
+ "grad_norm": 0.1656438261270523,
3050
+ "learning_rate": 6.439309443990532e-06,
3051
+ "loss": 0.0146,
3052
+ "step": 5070
3053
+ },
3054
+ {
3055
+ "grad_norm": 0.1890212744474411,
3056
+ "learning_rate": 6.304689333302416e-06,
3057
+ "loss": 0.0194,
3058
+ "step": 5080
3059
+ },
3060
+ {
3061
+ "grad_norm": 0.10458321869373322,
3062
+ "learning_rate": 6.171396567455051e-06,
3063
+ "loss": 0.0186,
3064
+ "step": 5090
3065
+ },
3066
+ {
3067
+ "grad_norm": 0.1596899926662445,
3068
+ "learning_rate": 6.039435195520604e-06,
3069
+ "loss": 0.0197,
3070
+ "step": 5100
3071
+ },
3072
+ {
3073
+ "grad_norm": 0.14461831748485565,
3074
+ "learning_rate": 5.908809226127054e-06,
3075
+ "loss": 0.0182,
3076
+ "step": 5110
3077
+ },
3078
+ {
3079
+ "grad_norm": 0.15653514862060547,
3080
+ "learning_rate": 5.779522627336537e-06,
3081
+ "loss": 0.0263,
3082
+ "step": 5120
3083
+ },
3084
+ {
3085
+ "grad_norm": 0.12199040502309799,
3086
+ "learning_rate": 5.651579326524709e-06,
3087
+ "loss": 0.0158,
3088
+ "step": 5130
3089
+ },
3090
+ {
3091
+ "grad_norm": 0.2367057204246521,
3092
+ "learning_rate": 5.524983210261481e-06,
3093
+ "loss": 0.0184,
3094
+ "step": 5140
3095
+ },
3096
+ {
3097
+ "grad_norm": 0.1951921284198761,
3098
+ "learning_rate": 5.399738124192988e-06,
3099
+ "loss": 0.0171,
3100
+ "step": 5150
3101
+ },
3102
+ {
3103
+ "grad_norm": 0.1722964495420456,
3104
+ "learning_rate": 5.2758478729247164e-06,
3105
+ "loss": 0.0203,
3106
+ "step": 5160
3107
+ },
3108
+ {
3109
+ "grad_norm": 0.15574219822883606,
3110
+ "learning_rate": 5.153316219905946e-06,
3111
+ "loss": 0.0128,
3112
+ "step": 5170
3113
+ },
3114
+ {
3115
+ "grad_norm": 0.16298823058605194,
3116
+ "learning_rate": 5.032146887315448e-06,
3117
+ "loss": 0.0167,
3118
+ "step": 5180
3119
+ },
3120
+ {
3121
+ "grad_norm": 0.19788341224193573,
3122
+ "learning_rate": 4.91234355594839e-06,
3123
+ "loss": 0.0198,
3124
+ "step": 5190
3125
+ },
3126
+ {
3127
+ "grad_norm": 0.1816922277212143,
3128
+ "learning_rate": 4.7939098651045235e-06,
3129
+ "loss": 0.0212,
3130
+ "step": 5200
3131
+ },
3132
+ {
3133
+ "grad_norm": 0.2591005265712738,
3134
+ "learning_rate": 4.67684941247768e-06,
3135
+ "loss": 0.0191,
3136
+ "step": 5210
3137
+ },
3138
+ {
3139
+ "grad_norm": 0.1402284801006317,
3140
+ "learning_rate": 4.5611657540464036e-06,
3141
+ "loss": 0.0181,
3142
+ "step": 5220
3143
+ },
3144
+ {
3145
+ "grad_norm": 0.17382720112800598,
3146
+ "learning_rate": 4.446862403965984e-06,
3147
+ "loss": 0.0184,
3148
+ "step": 5230
3149
+ },
3150
+ {
3151
+ "grad_norm": 0.1312701553106308,
3152
+ "learning_rate": 4.333942834461702e-06,
3153
+ "loss": 0.0153,
3154
+ "step": 5240
3155
+ },
3156
+ {
3157
+ "grad_norm": 0.20298051834106445,
3158
+ "learning_rate": 4.222410475723326e-06,
3159
+ "loss": 0.0158,
3160
+ "step": 5250
3161
+ },
3162
+ {
3163
+ "grad_norm": 0.10563281178474426,
3164
+ "learning_rate": 4.112268715800943e-06,
3165
+ "loss": 0.0196,
3166
+ "step": 5260
3167
+ },
3168
+ {
3169
+ "grad_norm": 0.15568211674690247,
3170
+ "learning_rate": 4.003520900502028e-06,
3171
+ "loss": 0.0189,
3172
+ "step": 5270
3173
+ },
3174
+ {
3175
+ "grad_norm": 0.15784397721290588,
3176
+ "learning_rate": 3.8961703332898e-06,
3177
+ "loss": 0.0184,
3178
+ "step": 5280
3179
+ },
3180
+ {
3181
+ "grad_norm": 0.1736660897731781,
3182
+ "learning_rate": 3.790220275182854e-06,
3183
+ "loss": 0.0214,
3184
+ "step": 5290
3185
+ },
3186
+ {
3187
+ "grad_norm": 0.19348296523094177,
3188
+ "learning_rate": 3.685673944656176e-06,
3189
+ "loss": 0.0175,
3190
+ "step": 5300
3191
+ },
3192
+ {
3193
+ "grad_norm": 0.17948998510837555,
3194
+ "learning_rate": 3.582534517543268e-06,
3195
+ "loss": 0.0166,
3196
+ "step": 5310
3197
+ },
3198
+ {
3199
+ "grad_norm": 0.24264298379421234,
3200
+ "learning_rate": 3.4808051269397512e-06,
3201
+ "loss": 0.0206,
3202
+ "step": 5320
3203
+ },
3204
+ {
3205
+ "grad_norm": 0.15983110666275024,
3206
+ "learning_rate": 3.380488863108183e-06,
3207
+ "loss": 0.0164,
3208
+ "step": 5330
3209
+ },
3210
+ {
3211
+ "grad_norm": 0.20391100645065308,
3212
+ "learning_rate": 3.2815887733841365e-06,
3213
+ "loss": 0.0224,
3214
+ "step": 5340
3215
+ },
3216
+ {
3217
+ "grad_norm": 0.18342769145965576,
3218
+ "learning_rate": 3.1841078620836683e-06,
3219
+ "loss": 0.0168,
3220
+ "step": 5350
3221
+ },
3222
+ {
3223
+ "grad_norm": 0.1805984526872635,
3224
+ "learning_rate": 3.0880490904120874e-06,
3225
+ "loss": 0.0205,
3226
+ "step": 5360
3227
+ },
3228
+ {
3229
+ "grad_norm": 0.18293620645999908,
3230
+ "learning_rate": 2.9934153763739205e-06,
3231
+ "loss": 0.0226,
3232
+ "step": 5370
3233
+ },
3234
+ {
3235
+ "grad_norm": 0.1255396008491516,
3236
+ "learning_rate": 2.9002095946843277e-06,
3237
+ "loss": 0.0164,
3238
+ "step": 5380
3239
+ },
3240
+ {
3241
+ "grad_norm": 0.13082446157932281,
3242
+ "learning_rate": 2.8084345766817676e-06,
3243
+ "loss": 0.0146,
3244
+ "step": 5390
3245
+ },
3246
+ {
3247
+ "grad_norm": 0.18769370019435883,
3248
+ "learning_rate": 2.718093110241976e-06,
3249
+ "loss": 0.017,
3250
+ "step": 5400
3251
+ },
3252
+ {
3253
+ "grad_norm": 0.13325811922550201,
3254
+ "learning_rate": 2.6291879396933004e-06,
3255
+ "loss": 0.017,
3256
+ "step": 5410
3257
+ },
3258
+ {
3259
+ "grad_norm": 0.1486356258392334,
3260
+ "learning_rate": 2.541721765733318e-06,
3261
+ "loss": 0.0136,
3262
+ "step": 5420
3263
+ },
3264
+ {
3265
+ "grad_norm": 0.22683610022068024,
3266
+ "learning_rate": 2.455697245346783e-06,
3267
+ "loss": 0.0185,
3268
+ "step": 5430
3269
+ },
3270
+ {
3271
+ "grad_norm": 0.14576876163482666,
3272
+ "learning_rate": 2.371116991724953e-06,
3273
+ "loss": 0.017,
3274
+ "step": 5440
3275
+ },
3276
+ {
3277
+ "grad_norm": 0.28364524245262146,
3278
+ "learning_rate": 2.2879835741861586e-06,
3279
+ "loss": 0.0224,
3280
+ "step": 5450
3281
+ },
3282
+ {
3283
+ "grad_norm": 0.1426982879638672,
3284
+ "learning_rate": 2.206299518097804e-06,
3285
+ "loss": 0.0148,
3286
+ "step": 5460
3287
+ },
3288
+ {
3289
+ "grad_norm": 0.14432327449321747,
3290
+ "learning_rate": 2.1260673047996227e-06,
3291
+ "loss": 0.0153,
3292
+ "step": 5470
3293
+ },
3294
+ {
3295
+ "grad_norm": 0.15404032170772552,
3296
+ "learning_rate": 2.047289371528299e-06,
3297
+ "loss": 0.018,
3298
+ "step": 5480
3299
+ },
3300
+ {
3301
+ "grad_norm": 0.16218610107898712,
3302
+ "learning_rate": 1.96996811134344e-06,
3303
+ "loss": 0.0126,
3304
+ "step": 5490
3305
+ },
3306
+ {
3307
+ "grad_norm": 0.14973685145378113,
3308
+ "learning_rate": 1.8941058730549132e-06,
3309
+ "loss": 0.014,
3310
+ "step": 5500
3311
+ },
3312
+ {
3313
+ "grad_norm": 0.16633783280849457,
3314
+ "learning_rate": 1.8197049611514194e-06,
3315
+ "loss": 0.0176,
3316
+ "step": 5510
3317
+ },
3318
+ {
3319
+ "grad_norm": 0.11517109721899033,
3320
+ "learning_rate": 1.7467676357305561e-06,
3321
+ "loss": 0.0141,
3322
+ "step": 5520
3323
+ },
3324
+ {
3325
+ "grad_norm": 0.18079155683517456,
3326
+ "learning_rate": 1.6752961124301415e-06,
3327
+ "loss": 0.0235,
3328
+ "step": 5530
3329
+ },
3330
+ {
3331
+ "grad_norm": 0.12392091751098633,
3332
+ "learning_rate": 1.6052925623609049e-06,
3333
+ "loss": 0.0171,
3334
+ "step": 5540
3335
+ },
3336
+ {
3337
+ "grad_norm": 0.14315803349018097,
3338
+ "learning_rate": 1.5367591120405256e-06,
3339
+ "loss": 0.0128,
3340
+ "step": 5550
3341
+ },
3342
+ {
3343
+ "grad_norm": 0.21578791737556458,
3344
+ "learning_rate": 1.4696978433290653e-06,
3345
+ "loss": 0.0159,
3346
+ "step": 5560
3347
+ },
3348
+ {
3349
+ "grad_norm": 0.20014861226081848,
3350
+ "learning_rate": 1.4041107933656928e-06,
3351
+ "loss": 0.0148,
3352
+ "step": 5570
3353
+ },
3354
+ {
3355
+ "grad_norm": 0.17368756234645844,
3356
+ "learning_rate": 1.339999954506821e-06,
3357
+ "loss": 0.0174,
3358
+ "step": 5580
3359
+ },
3360
+ {
3361
+ "grad_norm": 0.14039455354213715,
3362
+ "learning_rate": 1.2773672742655784e-06,
3363
+ "loss": 0.0165,
3364
+ "step": 5590
3365
+ },
3366
+ {
3367
+ "grad_norm": 0.22623248398303986,
3368
+ "learning_rate": 1.2162146552526399e-06,
3369
+ "loss": 0.0205,
3370
+ "step": 5600
3371
+ },
3372
+ {
3373
+ "grad_norm": 0.17202813923358917,
3374
+ "learning_rate": 1.1565439551184664e-06,
3375
+ "loss": 0.0179,
3376
+ "step": 5610
3377
+ },
3378
+ {
3379
+ "grad_norm": 0.19978828728199005,
3380
+ "learning_rate": 1.0983569864968346e-06,
3381
+ "loss": 0.0189,
3382
+ "step": 5620
3383
+ },
3384
+ {
3385
+ "grad_norm": 0.11077994853258133,
3386
+ "learning_rate": 1.0416555169497688e-06,
3387
+ "loss": 0.017,
3388
+ "step": 5630
3389
+ },
3390
+ {
3391
+ "grad_norm": 0.17590732872486115,
3392
+ "learning_rate": 9.864412689139123e-07,
3393
+ "loss": 0.0229,
3394
+ "step": 5640
3395
+ },
3396
+ {
3397
+ "grad_norm": 0.1582922339439392,
3398
+ "learning_rate": 9.327159196481138e-07,
3399
+ "loss": 0.011,
3400
+ "step": 5650
3401
+ },
3402
+ {
3403
+ "grad_norm": 0.16256555914878845,
3404
+ "learning_rate": 8.804811011825398e-07,
3405
+ "loss": 0.0188,
3406
+ "step": 5660
3407
+ },
3408
+ {
3409
+ "grad_norm": 0.1413237303495407,
3410
+ "learning_rate": 8.297384002690866e-07,
3411
+ "loss": 0.0163,
3412
+ "step": 5670
3413
+ },
3414
+ {
3415
+ "grad_norm": 0.18030285835266113,
3416
+ "learning_rate": 7.804893583331696e-07,
3417
+ "loss": 0.0156,
3418
+ "step": 5680
3419
+ },
3420
+ {
3421
+ "grad_norm": 0.1868339627981186,
3422
+ "learning_rate": 7.32735471426893e-07,
3423
+ "loss": 0.0199,
3424
+ "step": 5690
3425
+ },
3426
+ {
3427
+ "grad_norm": 0.1561463177204132,
3428
+ "learning_rate": 6.864781901836259e-07,
3429
+ "loss": 0.0145,
3430
+ "step": 5700
3431
+ },
3432
+ {
3433
+ "grad_norm": 0.1667407900094986,
3434
+ "learning_rate": 6.417189197739093e-07,
3435
+ "loss": 0.0192,
3436
+ "step": 5710
3437
+ },
3438
+ {
3439
+ "grad_norm": 0.19925780594348907,
3440
+ "learning_rate": 5.984590198627849e-07,
3441
+ "loss": 0.0199,
3442
+ "step": 5720
3443
+ },
3444
+ {
3445
+ "grad_norm": 0.19619081914424896,
3446
+ "learning_rate": 5.566998045685112e-07,
3447
+ "loss": 0.022,
3448
+ "step": 5730
3449
+ },
3450
+ {
3451
+ "grad_norm": 0.15238729119300842,
3452
+ "learning_rate": 5.164425424226016e-07,
3453
+ "loss": 0.0161,
3454
+ "step": 5740
3455
+ },
3456
+ {
3457
+ "grad_norm": 0.16857373714447021,
3458
+ "learning_rate": 4.776884563313266e-07,
3459
+ "loss": 0.0163,
3460
+ "step": 5750
3461
+ },
3462
+ {
3463
+ "grad_norm": 0.11748020350933075,
3464
+ "learning_rate": 4.404387235385443e-07,
3465
+ "loss": 0.0163,
3466
+ "step": 5760
3467
+ },
3468
+ {
3469
+ "grad_norm": 0.13694846630096436,
3470
+ "learning_rate": 4.0469447558995065e-07,
3471
+ "loss": 0.0181,
3472
+ "step": 5770
3473
+ },
3474
+ {
3475
+ "grad_norm": 0.1884409785270691,
3476
+ "learning_rate": 3.7045679829870175e-07,
3477
+ "loss": 0.0188,
3478
+ "step": 5780
3479
+ },
3480
+ {
3481
+ "grad_norm": 0.1570352464914322,
3482
+ "learning_rate": 3.377267317124233e-07,
3483
+ "loss": 0.0172,
3484
+ "step": 5790
3485
+ },
3486
+ {
3487
+ "grad_norm": 0.17494681477546692,
3488
+ "learning_rate": 3.0650527008162513e-07,
3489
+ "loss": 0.0165,
3490
+ "step": 5800
3491
+ },
3492
+ {
3493
+ "grad_norm": 0.1305803656578064,
3494
+ "learning_rate": 2.767933618295082e-07,
3495
+ "loss": 0.0144,
3496
+ "step": 5810
3497
+ },
3498
+ {
3499
+ "grad_norm": 0.19473201036453247,
3500
+ "learning_rate": 2.485919095231326e-07,
3501
+ "loss": 0.0156,
3502
+ "step": 5820
3503
+ },
3504
+ {
3505
+ "grad_norm": 0.1486295610666275,
3506
+ "learning_rate": 2.219017698460002e-07,
3507
+ "loss": 0.0201,
3508
+ "step": 5830
3509
+ },
3510
+ {
3511
+ "grad_norm": 0.19880466163158417,
3512
+ "learning_rate": 1.9672375357206452e-07,
3513
+ "loss": 0.0176,
3514
+ "step": 5840
3515
+ },
3516
+ {
3517
+ "grad_norm": 0.19250106811523438,
3518
+ "learning_rate": 1.73058625541056e-07,
3519
+ "loss": 0.0212,
3520
+ "step": 5850
3521
+ },
3522
+ {
3523
+ "grad_norm": 0.17275409400463104,
3524
+ "learning_rate": 1.5090710463527836e-07,
3525
+ "loss": 0.0164,
3526
+ "step": 5860
3527
+ },
3528
+ {
3529
+ "grad_norm": 0.11281181871891022,
3530
+ "learning_rate": 1.3026986375776485e-07,
3531
+ "loss": 0.0174,
3532
+ "step": 5870
3533
+ },
3534
+ {
3535
+ "grad_norm": 0.15450893342494965,
3536
+ "learning_rate": 1.1114752981183917e-07,
3537
+ "loss": 0.0247,
3538
+ "step": 5880
3539
+ },
3540
+ {
3541
+ "grad_norm": 0.15461589395999908,
3542
+ "learning_rate": 9.354068368204739e-08,
3543
+ "loss": 0.0153,
3544
+ "step": 5890
3545
+ },
3546
+ {
3547
+ "grad_norm": 0.15060506761074066,
3548
+ "learning_rate": 7.744986021656076e-08,
3549
+ "loss": 0.0182,
3550
+ "step": 5900
3551
+ },
3552
+ {
3553
+ "grad_norm": 0.12657040357589722,
3554
+ "learning_rate": 6.287554821087783e-08,
3555
+ "loss": 0.0173,
3556
+ "step": 5910
3557
+ },
3558
+ {
3559
+ "grad_norm": 0.1401786506175995,
3560
+ "learning_rate": 4.981819039300284e-08,
3561
+ "loss": 0.0192,
3562
+ "step": 5920
3563
+ },
3564
+ {
3565
+ "grad_norm": 0.15238840878009796,
3566
+ "learning_rate": 3.827818341000655e-08,
3567
+ "loss": 0.0157,
3568
+ "step": 5930
3569
+ },
3570
+ {
3571
+ "grad_norm": 0.10546451807022095,
3572
+ "learning_rate": 2.8255877815946963e-08,
3573
+ "loss": 0.0147,
3574
+ "step": 5940
3575
+ },
3576
+ {
3577
+ "grad_norm": 0.15890178084373474,
3578
+ "learning_rate": 1.9751578061244504e-08,
3579
+ "loss": 0.0149,
3580
+ "step": 5950
3581
+ },
3582
+ {
3583
+ "grad_norm": 0.17914965748786926,
3584
+ "learning_rate": 1.2765542483417214e-08,
3585
+ "loss": 0.0151,
3586
+ "step": 5960
3587
+ },
3588
+ {
3589
+ "grad_norm": 0.20467838644981384,
3590
+ "learning_rate": 7.2979832992592365e-09,
3591
+ "loss": 0.0157,
3592
+ "step": 5970
3593
+ },
3594
+ {
3595
+ "grad_norm": 0.21326901018619537,
3596
+ "learning_rate": 3.349066598362649e-09,
3597
+ "loss": 0.0195,
3598
+ "step": 5980
3599
+ },
3600
+ {
3601
+ "grad_norm": 0.20198704302310944,
3602
+ "learning_rate": 9.189123380826114e-10,
3603
+ "loss": 0.0161,
3604
+ "step": 5990
3605
+ },
3606
+ {
3607
+ "grad_norm": 0.1654694527387619,
3608
+ "learning_rate": 7.594339912486703e-12,
3609
+ "loss": 0.0169,
3610
+ "step": 6000
3611
+ }
3612
+ ],
3613
+ "logging_steps": 10,
3614
+ "max_steps": 6000,
3615
+ "num_input_tokens_seen": 0,
3616
+ "num_train_epochs": 9223372036854775807,
3617
+ "save_steps": 2000,
3618
+ "stateful_callbacks": {
3619
+ "TrainerControl": {
3620
+ "args": {
3621
+ "should_epoch_stop": false,
3622
+ "should_evaluate": false,
3623
+ "should_log": false,
3624
+ "should_save": true,
3625
+ "should_training_stop": true
3626
+ },
3627
+ "attributes": {}
3628
+ }
3629
+ },
3630
+ "total_flos": 0.0,
3631
+ "train_batch_size": 32,
3632
+ "trial_name": null,
3633
+ "trial_params": null
3634
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d06dde1b8311531971279bb8ba6e785393d2e02e5034076d7b5d6e9c0dd8990a
3
+ size 5777
wandb_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"project": "finetune-gr00t-n1d7", "run_id": "6f4f0bf8-fd17-47a5-a3c5-6bccc5bd08f8"}