upload rmu_Llama-3.1-8B_20260427_203908
Browse files- .gitattributes +4 -0
- checkpoint-200/config.json +36 -0
- checkpoint-200/generation_config.json +12 -0
- checkpoint-200/model.safetensors +3 -0
- checkpoint-200/optimizer.pt +3 -0
- checkpoint-200/rng_state.pth +3 -0
- checkpoint-200/scheduler.pt +3 -0
- checkpoint-200/tokenizer.json +3 -0
- checkpoint-200/tokenizer_config.json +15 -0
- checkpoint-200/trainer_state.json +661 -0
- checkpoint-200/training_args.bin +3 -0
- checkpoint-400/config.json +36 -0
- checkpoint-400/generation_config.json +12 -0
- checkpoint-400/model.safetensors +3 -0
- checkpoint-400/optimizer.pt +3 -0
- checkpoint-400/rng_state.pth +3 -0
- checkpoint-400/scheduler.pt +3 -0
- checkpoint-400/tokenizer.json +3 -0
- checkpoint-400/tokenizer_config.json +15 -0
- checkpoint-400/trainer_state.json +1281 -0
- checkpoint-400/training_args.bin +3 -0
- checkpoint-534/config.json +36 -0
- checkpoint-534/generation_config.json +12 -0
- checkpoint-534/model.safetensors +3 -0
- checkpoint-534/optimizer.pt +3 -0
- checkpoint-534/rng_state.pth +3 -0
- checkpoint-534/scheduler.pt +3 -0
- checkpoint-534/tokenizer.json +3 -0
- checkpoint-534/tokenizer_config.json +15 -0
- checkpoint-534/trainer_state.json +1708 -0
- checkpoint-534/training_args.bin +3 -0
- config.json +36 -0
- config.yaml +35 -0
- generation_config.json +12 -0
- model.safetensors +3 -0
- tokenizer.json +3 -0
- tokenizer_config.json +15 -0
- train.log +7 -0
- train_summary.json +48 -0
- training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,7 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
checkpoint-200/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 37 |
+
checkpoint-400/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 38 |
+
checkpoint-534/tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
| 39 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
checkpoint-200/config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"LlamaForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 128000,
|
| 8 |
+
"dtype": "bfloat16",
|
| 9 |
+
"eos_token_id": 128001,
|
| 10 |
+
"head_dim": 128,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 4096,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 14336,
|
| 15 |
+
"max_position_embeddings": 131072,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 32,
|
| 19 |
+
"num_hidden_layers": 32,
|
| 20 |
+
"num_key_value_heads": 8,
|
| 21 |
+
"pad_token_id": 128001,
|
| 22 |
+
"pretraining_tp": 1,
|
| 23 |
+
"rms_norm_eps": 1e-05,
|
| 24 |
+
"rope_parameters": {
|
| 25 |
+
"factor": 8.0,
|
| 26 |
+
"high_freq_factor": 4.0,
|
| 27 |
+
"low_freq_factor": 1.0,
|
| 28 |
+
"original_max_position_embeddings": 8192,
|
| 29 |
+
"rope_theta": 500000.0,
|
| 30 |
+
"rope_type": "llama3"
|
| 31 |
+
},
|
| 32 |
+
"tie_word_embeddings": false,
|
| 33 |
+
"transformers_version": "5.6.0",
|
| 34 |
+
"use_cache": false,
|
| 35 |
+
"vocab_size": 128256
|
| 36 |
+
}
|
checkpoint-200/generation_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 128000,
|
| 4 |
+
"do_sample": true,
|
| 5 |
+
"eos_token_id": [
|
| 6 |
+
128001
|
| 7 |
+
],
|
| 8 |
+
"pad_token_id": 128001,
|
| 9 |
+
"temperature": 0.6,
|
| 10 |
+
"top_p": 0.9,
|
| 11 |
+
"transformers_version": "5.6.0"
|
| 12 |
+
}
|
checkpoint-200/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e07e5b561ec0fde5a2ea146fd143d76eb8e9d222941d37d51150074159c21e9
|
| 3 |
+
size 16060556616
|
checkpoint-200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ce4a1c70427daa6eb23ce55c4b25a94a7aa6674844d916890e6ccff2ce738e36
|
| 3 |
+
size 9080995015
|
checkpoint-200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:91317c5e4abc67d24eaa48e06aae01d8e24c1621d2d7b287b5317f05ead1b119
|
| 3 |
+
size 14645
|
checkpoint-200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6064554228f0a8789adc25287f0e1251cf8bf5c5613edc79a0122720bfc71c3e
|
| 3 |
+
size 1465
|
checkpoint-200/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
|
| 3 |
+
size 17209920
|
checkpoint-200/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "<|begin_of_text|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "<|end_of_text|>",
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"local_files_only": false,
|
| 8 |
+
"model_input_names": [
|
| 9 |
+
"input_ids",
|
| 10 |
+
"attention_mask"
|
| 11 |
+
],
|
| 12 |
+
"model_max_length": 131072,
|
| 13 |
+
"pad_token": "<|end_of_text|>",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend"
|
| 15 |
+
}
|
checkpoint-200/trainer_state.json
ADDED
|
@@ -0,0 +1,661 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.7511737089201878,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 200,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0,
|
| 14 |
+
"forget_loss": 0.00946044921875,
|
| 15 |
+
"retain_loss": 0.0,
|
| 16 |
+
"step": 0
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0,
|
| 20 |
+
"forget_loss": 0.00994873046875,
|
| 21 |
+
"retain_loss": 0.0,
|
| 22 |
+
"step": 0
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 0,
|
| 26 |
+
"forget_loss": 0.00958251953125,
|
| 27 |
+
"retain_loss": 0.0,
|
| 28 |
+
"step": 0
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0,
|
| 32 |
+
"forget_loss": 0.009521484375,
|
| 33 |
+
"retain_loss": 0.0,
|
| 34 |
+
"step": 0
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.003755868544600939,
|
| 38 |
+
"grad_norm": 0.11474609375,
|
| 39 |
+
"learning_rate": 0.0,
|
| 40 |
+
"loss": 0.0096282958984375,
|
| 41 |
+
"step": 1
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.03755868544600939,
|
| 45 |
+
"grad_norm": 0.09814453125,
|
| 46 |
+
"learning_rate": 4.5e-06,
|
| 47 |
+
"loss": 0.009236653645833334,
|
| 48 |
+
"step": 10
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.03755868544600939,
|
| 52 |
+
"forget_loss": 0.00799560546875,
|
| 53 |
+
"retain_loss": 1.919269561767578e-05,
|
| 54 |
+
"step": 10
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.03755868544600939,
|
| 58 |
+
"forget_loss": 0.008056640625,
|
| 59 |
+
"retain_loss": 1.6570091247558594e-05,
|
| 60 |
+
"step": 10
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.03755868544600939,
|
| 64 |
+
"forget_loss": 0.0081787109375,
|
| 65 |
+
"retain_loss": 2.2649765014648438e-05,
|
| 66 |
+
"step": 10
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.03755868544600939,
|
| 70 |
+
"forget_loss": 0.00823974609375,
|
| 71 |
+
"retain_loss": 2.1457672119140625e-05,
|
| 72 |
+
"step": 10
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07511737089201878,
|
| 76 |
+
"grad_norm": 0.047607421875,
|
| 77 |
+
"learning_rate": 9.5e-06,
|
| 78 |
+
"loss": 0.00635223388671875,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07511737089201878,
|
| 83 |
+
"forget_loss": 0.003662109375,
|
| 84 |
+
"retain_loss": 0.00022602081298828125,
|
| 85 |
+
"step": 20
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.07511737089201878,
|
| 89 |
+
"forget_loss": 0.004425048828125,
|
| 90 |
+
"retain_loss": 0.0002880096435546875,
|
| 91 |
+
"step": 20
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.07511737089201878,
|
| 95 |
+
"forget_loss": 0.0037078857421875,
|
| 96 |
+
"retain_loss": 0.00020885467529296875,
|
| 97 |
+
"step": 20
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.07511737089201878,
|
| 101 |
+
"forget_loss": 0.00421142578125,
|
| 102 |
+
"retain_loss": 0.00021457672119140625,
|
| 103 |
+
"step": 20
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.11267605633802817,
|
| 107 |
+
"grad_norm": 0.123046875,
|
| 108 |
+
"learning_rate": 9.82490272373541e-06,
|
| 109 |
+
"loss": 0.00320587158203125,
|
| 110 |
+
"step": 30
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.11267605633802817,
|
| 114 |
+
"forget_loss": 0.0018310546875,
|
| 115 |
+
"retain_loss": 0.00054168701171875,
|
| 116 |
+
"step": 30
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.11267605633802817,
|
| 120 |
+
"forget_loss": 0.00177001953125,
|
| 121 |
+
"retain_loss": 0.0004177093505859375,
|
| 122 |
+
"step": 30
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.11267605633802817,
|
| 126 |
+
"forget_loss": 0.00177001953125,
|
| 127 |
+
"retain_loss": 0.0002956390380859375,
|
| 128 |
+
"step": 30
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.11267605633802817,
|
| 132 |
+
"forget_loss": 0.0016937255859375,
|
| 133 |
+
"retain_loss": 0.000331878662109375,
|
| 134 |
+
"step": 30
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.15023474178403756,
|
| 138 |
+
"grad_norm": 0.05517578125,
|
| 139 |
+
"learning_rate": 9.63035019455253e-06,
|
| 140 |
+
"loss": 0.0016145706176757812,
|
| 141 |
+
"step": 40
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.15023474178403756,
|
| 145 |
+
"forget_loss": 0.0007781982421875,
|
| 146 |
+
"retain_loss": 0.00021648406982421875,
|
| 147 |
+
"step": 40
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.15023474178403756,
|
| 151 |
+
"forget_loss": 0.000762939453125,
|
| 152 |
+
"retain_loss": 0.0004100799560546875,
|
| 153 |
+
"step": 40
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.15023474178403756,
|
| 157 |
+
"forget_loss": 0.000766754150390625,
|
| 158 |
+
"retain_loss": 0.000331878662109375,
|
| 159 |
+
"step": 40
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.15023474178403756,
|
| 163 |
+
"forget_loss": 0.0007781982421875,
|
| 164 |
+
"retain_loss": 0.00038909912109375,
|
| 165 |
+
"step": 40
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.18779342723004694,
|
| 169 |
+
"grad_norm": 0.1416015625,
|
| 170 |
+
"learning_rate": 9.435797665369651e-06,
|
| 171 |
+
"loss": 0.000954437255859375,
|
| 172 |
+
"step": 50
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.18779342723004694,
|
| 176 |
+
"forget_loss": 0.0003452301025390625,
|
| 177 |
+
"retain_loss": 0.00020599365234375,
|
| 178 |
+
"step": 50
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.18779342723004694,
|
| 182 |
+
"forget_loss": 0.0003814697265625,
|
| 183 |
+
"retain_loss": 0.00020694732666015625,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.18779342723004694,
|
| 188 |
+
"forget_loss": 0.000415802001953125,
|
| 189 |
+
"retain_loss": 0.0003604888916015625,
|
| 190 |
+
"step": 50
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18779342723004694,
|
| 194 |
+
"forget_loss": 0.0003719329833984375,
|
| 195 |
+
"retain_loss": 0.0001964569091796875,
|
| 196 |
+
"step": 50
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.22535211267605634,
|
| 200 |
+
"grad_norm": 0.06396484375,
|
| 201 |
+
"learning_rate": 9.241245136186772e-06,
|
| 202 |
+
"loss": 0.0008329868316650391,
|
| 203 |
+
"step": 60
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.22535211267605634,
|
| 207 |
+
"forget_loss": 0.0002574920654296875,
|
| 208 |
+
"retain_loss": 0.0002727508544921875,
|
| 209 |
+
"step": 60
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.22535211267605634,
|
| 213 |
+
"forget_loss": 0.00019931793212890625,
|
| 214 |
+
"retain_loss": 0.00010919570922851562,
|
| 215 |
+
"step": 60
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.22535211267605634,
|
| 219 |
+
"forget_loss": 0.0001983642578125,
|
| 220 |
+
"retain_loss": 0.000202178955078125,
|
| 221 |
+
"step": 60
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.22535211267605634,
|
| 225 |
+
"forget_loss": 0.00023365020751953125,
|
| 226 |
+
"retain_loss": 0.0002002716064453125,
|
| 227 |
+
"step": 60
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.26291079812206575,
|
| 231 |
+
"grad_norm": 0.28515625,
|
| 232 |
+
"learning_rate": 9.046692607003892e-06,
|
| 233 |
+
"loss": 0.0006081581115722657,
|
| 234 |
+
"step": 70
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.26291079812206575,
|
| 238 |
+
"forget_loss": 0.00018405914306640625,
|
| 239 |
+
"retain_loss": 0.0002689361572265625,
|
| 240 |
+
"step": 70
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.26291079812206575,
|
| 244 |
+
"forget_loss": 0.0001811981201171875,
|
| 245 |
+
"retain_loss": 8.535385131835938e-05,
|
| 246 |
+
"step": 70
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.26291079812206575,
|
| 250 |
+
"forget_loss": 0.00022792816162109375,
|
| 251 |
+
"retain_loss": 0.00010728836059570312,
|
| 252 |
+
"step": 70
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.26291079812206575,
|
| 256 |
+
"forget_loss": 0.0003643035888671875,
|
| 257 |
+
"retain_loss": 0.00011777877807617188,
|
| 258 |
+
"step": 70
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.3004694835680751,
|
| 262 |
+
"grad_norm": 0.07177734375,
|
| 263 |
+
"learning_rate": 8.852140077821012e-06,
|
| 264 |
+
"loss": 0.0004550933837890625,
|
| 265 |
+
"step": 80
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.3004694835680751,
|
| 269 |
+
"forget_loss": 0.0002155303955078125,
|
| 270 |
+
"retain_loss": 0.000110626220703125,
|
| 271 |
+
"step": 80
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.3004694835680751,
|
| 275 |
+
"forget_loss": 0.00014019012451171875,
|
| 276 |
+
"retain_loss": 6.723403930664062e-05,
|
| 277 |
+
"step": 80
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.3004694835680751,
|
| 281 |
+
"forget_loss": 0.0001773834228515625,
|
| 282 |
+
"retain_loss": 0.000331878662109375,
|
| 283 |
+
"step": 80
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3004694835680751,
|
| 287 |
+
"forget_loss": 0.00016498565673828125,
|
| 288 |
+
"retain_loss": 9.489059448242188e-05,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3380281690140845,
|
| 293 |
+
"grad_norm": 0.0810546875,
|
| 294 |
+
"learning_rate": 8.657587548638134e-06,
|
| 295 |
+
"loss": 0.0003980875015258789,
|
| 296 |
+
"step": 90
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3380281690140845,
|
| 300 |
+
"forget_loss": 0.00015735626220703125,
|
| 301 |
+
"retain_loss": 0.0001544952392578125,
|
| 302 |
+
"step": 90
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.3380281690140845,
|
| 306 |
+
"forget_loss": 0.000133514404296875,
|
| 307 |
+
"retain_loss": 0.00080108642578125,
|
| 308 |
+
"step": 90
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.3380281690140845,
|
| 312 |
+
"forget_loss": 0.00014781951904296875,
|
| 313 |
+
"retain_loss": 0.000431060791015625,
|
| 314 |
+
"step": 90
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.3380281690140845,
|
| 318 |
+
"forget_loss": 0.00014400482177734375,
|
| 319 |
+
"retain_loss": 0.00031280517578125,
|
| 320 |
+
"step": 90
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.3755868544600939,
|
| 324 |
+
"grad_norm": 0.025146484375,
|
| 325 |
+
"learning_rate": 8.463035019455253e-06,
|
| 326 |
+
"loss": 0.00032124519348144533,
|
| 327 |
+
"step": 100
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.3755868544600939,
|
| 331 |
+
"forget_loss": 0.00015163421630859375,
|
| 332 |
+
"retain_loss": 0.00015926361083984375,
|
| 333 |
+
"step": 100
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.3755868544600939,
|
| 337 |
+
"forget_loss": 0.0001506805419921875,
|
| 338 |
+
"retain_loss": 0.00011587142944335938,
|
| 339 |
+
"step": 100
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3755868544600939,
|
| 343 |
+
"forget_loss": 0.00013065338134765625,
|
| 344 |
+
"retain_loss": 0.000125885009765625,
|
| 345 |
+
"step": 100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3755868544600939,
|
| 349 |
+
"forget_loss": 0.0001239776611328125,
|
| 350 |
+
"retain_loss": 0.0001277923583984375,
|
| 351 |
+
"step": 100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.4131455399061033,
|
| 355 |
+
"grad_norm": 0.06884765625,
|
| 356 |
+
"learning_rate": 8.268482490272373e-06,
|
| 357 |
+
"loss": 0.00034072399139404295,
|
| 358 |
+
"step": 110
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.4131455399061033,
|
| 362 |
+
"forget_loss": 0.0001544952392578125,
|
| 363 |
+
"retain_loss": 7.2479248046875e-05,
|
| 364 |
+
"step": 110
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.4131455399061033,
|
| 368 |
+
"forget_loss": 0.0001373291015625,
|
| 369 |
+
"retain_loss": 0.000270843505859375,
|
| 370 |
+
"step": 110
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.4131455399061033,
|
| 374 |
+
"forget_loss": 0.0001125335693359375,
|
| 375 |
+
"retain_loss": 9.965896606445312e-05,
|
| 376 |
+
"step": 110
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.4131455399061033,
|
| 380 |
+
"forget_loss": 0.00011205673217773438,
|
| 381 |
+
"retain_loss": 6.866455078125e-05,
|
| 382 |
+
"step": 110
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.4507042253521127,
|
| 386 |
+
"grad_norm": 0.08544921875,
|
| 387 |
+
"learning_rate": 8.073929961089495e-06,
|
| 388 |
+
"loss": 0.0002372264862060547,
|
| 389 |
+
"step": 120
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.4507042253521127,
|
| 393 |
+
"forget_loss": 0.000179290771484375,
|
| 394 |
+
"retain_loss": 7.05718994140625e-05,
|
| 395 |
+
"step": 120
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.4507042253521127,
|
| 399 |
+
"forget_loss": 0.0001277923583984375,
|
| 400 |
+
"retain_loss": 9.250640869140625e-05,
|
| 401 |
+
"step": 120
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4507042253521127,
|
| 405 |
+
"forget_loss": 0.00013446807861328125,
|
| 406 |
+
"retain_loss": 8.630752563476562e-05,
|
| 407 |
+
"step": 120
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.4507042253521127,
|
| 411 |
+
"forget_loss": 0.0001735687255859375,
|
| 412 |
+
"retain_loss": 8.106231689453125e-05,
|
| 413 |
+
"step": 120
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.48826291079812206,
|
| 417 |
+
"grad_norm": 0.0361328125,
|
| 418 |
+
"learning_rate": 7.879377431906614e-06,
|
| 419 |
+
"loss": 0.0002105236053466797,
|
| 420 |
+
"step": 130
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.48826291079812206,
|
| 424 |
+
"forget_loss": 0.00010776519775390625,
|
| 425 |
+
"retain_loss": 5.221366882324219e-05,
|
| 426 |
+
"step": 130
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.48826291079812206,
|
| 430 |
+
"forget_loss": 0.00010728836059570312,
|
| 431 |
+
"retain_loss": 5.1021575927734375e-05,
|
| 432 |
+
"step": 130
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.48826291079812206,
|
| 436 |
+
"forget_loss": 0.00010204315185546875,
|
| 437 |
+
"retain_loss": 0.00011110305786132812,
|
| 438 |
+
"step": 130
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.48826291079812206,
|
| 442 |
+
"forget_loss": 0.000156402587890625,
|
| 443 |
+
"retain_loss": 7.05718994140625e-05,
|
| 444 |
+
"step": 130
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5258215962441315,
|
| 448 |
+
"grad_norm": 0.035400390625,
|
| 449 |
+
"learning_rate": 7.684824902723736e-06,
|
| 450 |
+
"loss": 0.000272369384765625,
|
| 451 |
+
"step": 140
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5258215962441315,
|
| 455 |
+
"forget_loss": 0.0001087188720703125,
|
| 456 |
+
"retain_loss": 7.009506225585938e-05,
|
| 457 |
+
"step": 140
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5258215962441315,
|
| 461 |
+
"forget_loss": 0.0001239776611328125,
|
| 462 |
+
"retain_loss": 7.963180541992188e-05,
|
| 463 |
+
"step": 140
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.5258215962441315,
|
| 467 |
+
"forget_loss": 0.00017070770263671875,
|
| 468 |
+
"retain_loss": 6.29425048828125e-05,
|
| 469 |
+
"step": 140
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.5258215962441315,
|
| 473 |
+
"forget_loss": 0.00013446807861328125,
|
| 474 |
+
"retain_loss": 5.5789947509765625e-05,
|
| 475 |
+
"step": 140
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.5633802816901409,
|
| 479 |
+
"grad_norm": 0.0177001953125,
|
| 480 |
+
"learning_rate": 7.4902723735408565e-06,
|
| 481 |
+
"loss": 0.00019230842590332032,
|
| 482 |
+
"step": 150
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.5633802816901409,
|
| 486 |
+
"forget_loss": 0.00010585784912109375,
|
| 487 |
+
"retain_loss": 5.793571472167969e-05,
|
| 488 |
+
"step": 150
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.5633802816901409,
|
| 492 |
+
"forget_loss": 0.00011157989501953125,
|
| 493 |
+
"retain_loss": 0.00015163421630859375,
|
| 494 |
+
"step": 150
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.5633802816901409,
|
| 498 |
+
"forget_loss": 8.821487426757812e-05,
|
| 499 |
+
"retain_loss": 6.866455078125e-05,
|
| 500 |
+
"step": 150
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5633802816901409,
|
| 504 |
+
"forget_loss": 9.059906005859375e-05,
|
| 505 |
+
"retain_loss": 0.00010251998901367188,
|
| 506 |
+
"step": 150
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.6009389671361502,
|
| 510 |
+
"grad_norm": 0.06591796875,
|
| 511 |
+
"learning_rate": 7.295719844357977e-06,
|
| 512 |
+
"loss": 0.00020923614501953126,
|
| 513 |
+
"step": 160
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.6009389671361502,
|
| 517 |
+
"forget_loss": 0.00010156631469726562,
|
| 518 |
+
"retain_loss": 8.535385131835938e-05,
|
| 519 |
+
"step": 160
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.6009389671361502,
|
| 523 |
+
"forget_loss": 9.441375732421875e-05,
|
| 524 |
+
"retain_loss": 7.05718994140625e-05,
|
| 525 |
+
"step": 160
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.6009389671361502,
|
| 529 |
+
"forget_loss": 9.489059448242188e-05,
|
| 530 |
+
"retain_loss": 6.151199340820312e-05,
|
| 531 |
+
"step": 160
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.6009389671361502,
|
| 535 |
+
"forget_loss": 9.679794311523438e-05,
|
| 536 |
+
"retain_loss": 5.984306335449219e-05,
|
| 537 |
+
"step": 160
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.6384976525821596,
|
| 541 |
+
"grad_norm": 0.0152587890625,
|
| 542 |
+
"learning_rate": 7.101167315175098e-06,
|
| 543 |
+
"loss": 0.0001931428909301758,
|
| 544 |
+
"step": 170
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.6384976525821596,
|
| 548 |
+
"forget_loss": 9.489059448242188e-05,
|
| 549 |
+
"retain_loss": 6.079673767089844e-05,
|
| 550 |
+
"step": 170
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.6384976525821596,
|
| 554 |
+
"forget_loss": 0.000110626220703125,
|
| 555 |
+
"retain_loss": 7.104873657226562e-05,
|
| 556 |
+
"step": 170
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6384976525821596,
|
| 560 |
+
"forget_loss": 0.00010776519775390625,
|
| 561 |
+
"retain_loss": 0.00011539459228515625,
|
| 562 |
+
"step": 170
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6384976525821596,
|
| 566 |
+
"forget_loss": 9.441375732421875e-05,
|
| 567 |
+
"retain_loss": 0.00011682510375976562,
|
| 568 |
+
"step": 170
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.676056338028169,
|
| 572 |
+
"grad_norm": 0.0306396484375,
|
| 573 |
+
"learning_rate": 6.9066147859922184e-06,
|
| 574 |
+
"loss": 0.00021104812622070314,
|
| 575 |
+
"step": 180
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.676056338028169,
|
| 579 |
+
"forget_loss": 0.000125885009765625,
|
| 580 |
+
"retain_loss": 0.0001239776611328125,
|
| 581 |
+
"step": 180
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.676056338028169,
|
| 585 |
+
"forget_loss": 0.00010395050048828125,
|
| 586 |
+
"retain_loss": 8.96453857421875e-05,
|
| 587 |
+
"step": 180
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 0.676056338028169,
|
| 591 |
+
"forget_loss": 8.869171142578125e-05,
|
| 592 |
+
"retain_loss": 8.20159912109375e-05,
|
| 593 |
+
"step": 180
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.676056338028169,
|
| 597 |
+
"forget_loss": 9.107589721679688e-05,
|
| 598 |
+
"retain_loss": 5.316734313964844e-05,
|
| 599 |
+
"step": 180
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.7136150234741784,
|
| 603 |
+
"grad_norm": 0.0230712890625,
|
| 604 |
+
"learning_rate": 6.712062256809338e-06,
|
| 605 |
+
"loss": 0.0001739978790283203,
|
| 606 |
+
"step": 190
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.7136150234741784,
|
| 610 |
+
"forget_loss": 9.059906005859375e-05,
|
| 611 |
+
"retain_loss": 5.793571472167969e-05,
|
| 612 |
+
"step": 190
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7136150234741784,
|
| 616 |
+
"forget_loss": 9.5367431640625e-05,
|
| 617 |
+
"retain_loss": 5.0067901611328125e-05,
|
| 618 |
+
"step": 190
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7136150234741784,
|
| 622 |
+
"forget_loss": 8.869171142578125e-05,
|
| 623 |
+
"retain_loss": 5.7220458984375e-05,
|
| 624 |
+
"step": 190
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.7136150234741784,
|
| 628 |
+
"forget_loss": 0.00010919570922851562,
|
| 629 |
+
"retain_loss": 4.9591064453125e-05,
|
| 630 |
+
"step": 190
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.7511737089201878,
|
| 634 |
+
"grad_norm": 0.01068115234375,
|
| 635 |
+
"learning_rate": 6.51750972762646e-06,
|
| 636 |
+
"loss": 0.00019364356994628907,
|
| 637 |
+
"step": 200
|
| 638 |
+
}
|
| 639 |
+
],
|
| 640 |
+
"logging_steps": 10,
|
| 641 |
+
"max_steps": 534,
|
| 642 |
+
"num_input_tokens_seen": 0,
|
| 643 |
+
"num_train_epochs": 2,
|
| 644 |
+
"save_steps": 200,
|
| 645 |
+
"stateful_callbacks": {
|
| 646 |
+
"TrainerControl": {
|
| 647 |
+
"args": {
|
| 648 |
+
"should_epoch_stop": false,
|
| 649 |
+
"should_evaluate": false,
|
| 650 |
+
"should_log": false,
|
| 651 |
+
"should_save": true,
|
| 652 |
+
"should_training_stop": false
|
| 653 |
+
},
|
| 654 |
+
"attributes": {}
|
| 655 |
+
}
|
| 656 |
+
},
|
| 657 |
+
"total_flos": 0.0,
|
| 658 |
+
"train_batch_size": 8,
|
| 659 |
+
"trial_name": null,
|
| 660 |
+
"trial_params": null
|
| 661 |
+
}
|
checkpoint-200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cde3986b5d2902c3143ac3a5311b9443a600348ae4fbcaab8a20b0760c034ff1
|
| 3 |
+
size 5265
|
checkpoint-400/config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"LlamaForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 128000,
|
| 8 |
+
"dtype": "bfloat16",
|
| 9 |
+
"eos_token_id": 128001,
|
| 10 |
+
"head_dim": 128,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 4096,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 14336,
|
| 15 |
+
"max_position_embeddings": 131072,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 32,
|
| 19 |
+
"num_hidden_layers": 32,
|
| 20 |
+
"num_key_value_heads": 8,
|
| 21 |
+
"pad_token_id": 128001,
|
| 22 |
+
"pretraining_tp": 1,
|
| 23 |
+
"rms_norm_eps": 1e-05,
|
| 24 |
+
"rope_parameters": {
|
| 25 |
+
"factor": 8.0,
|
| 26 |
+
"high_freq_factor": 4.0,
|
| 27 |
+
"low_freq_factor": 1.0,
|
| 28 |
+
"original_max_position_embeddings": 8192,
|
| 29 |
+
"rope_theta": 500000.0,
|
| 30 |
+
"rope_type": "llama3"
|
| 31 |
+
},
|
| 32 |
+
"tie_word_embeddings": false,
|
| 33 |
+
"transformers_version": "5.6.0",
|
| 34 |
+
"use_cache": false,
|
| 35 |
+
"vocab_size": 128256
|
| 36 |
+
}
|
checkpoint-400/generation_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 128000,
|
| 4 |
+
"do_sample": true,
|
| 5 |
+
"eos_token_id": [
|
| 6 |
+
128001
|
| 7 |
+
],
|
| 8 |
+
"pad_token_id": 128001,
|
| 9 |
+
"temperature": 0.6,
|
| 10 |
+
"top_p": 0.9,
|
| 11 |
+
"transformers_version": "5.6.0"
|
| 12 |
+
}
|
checkpoint-400/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd76307f41cce10999e3bc86c75ed43f40fd79e768624b63b74cd819586d6efc
|
| 3 |
+
size 16060556616
|
checkpoint-400/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9ed784a2bfbaad6e9388832dc6780e4bb5dd3dadc965329f60cc7e403208eb2a
|
| 3 |
+
size 9080995015
|
checkpoint-400/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6cb8976bef1b2d43f5c801e4f98bb4fe61d7eb52d27706beca6e7cb4cc9e590b
|
| 3 |
+
size 14645
|
checkpoint-400/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3105c86b7261131bcf69536aefea946cb903a5a7d2f97d1b1105ad908ac15e1b
|
| 3 |
+
size 1465
|
checkpoint-400/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
|
| 3 |
+
size 17209920
|
checkpoint-400/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "<|begin_of_text|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "<|end_of_text|>",
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"local_files_only": false,
|
| 8 |
+
"model_input_names": [
|
| 9 |
+
"input_ids",
|
| 10 |
+
"attention_mask"
|
| 11 |
+
],
|
| 12 |
+
"model_max_length": 131072,
|
| 13 |
+
"pad_token": "<|end_of_text|>",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend"
|
| 15 |
+
}
|
checkpoint-400/trainer_state.json
ADDED
|
@@ -0,0 +1,1281 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.4995305164319248,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 400,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0,
|
| 14 |
+
"forget_loss": 0.00946044921875,
|
| 15 |
+
"retain_loss": 0.0,
|
| 16 |
+
"step": 0
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0,
|
| 20 |
+
"forget_loss": 0.00994873046875,
|
| 21 |
+
"retain_loss": 0.0,
|
| 22 |
+
"step": 0
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 0,
|
| 26 |
+
"forget_loss": 0.00958251953125,
|
| 27 |
+
"retain_loss": 0.0,
|
| 28 |
+
"step": 0
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0,
|
| 32 |
+
"forget_loss": 0.009521484375,
|
| 33 |
+
"retain_loss": 0.0,
|
| 34 |
+
"step": 0
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.003755868544600939,
|
| 38 |
+
"grad_norm": 0.11474609375,
|
| 39 |
+
"learning_rate": 0.0,
|
| 40 |
+
"loss": 0.0096282958984375,
|
| 41 |
+
"step": 1
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.03755868544600939,
|
| 45 |
+
"grad_norm": 0.09814453125,
|
| 46 |
+
"learning_rate": 4.5e-06,
|
| 47 |
+
"loss": 0.009236653645833334,
|
| 48 |
+
"step": 10
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.03755868544600939,
|
| 52 |
+
"forget_loss": 0.00799560546875,
|
| 53 |
+
"retain_loss": 1.919269561767578e-05,
|
| 54 |
+
"step": 10
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.03755868544600939,
|
| 58 |
+
"forget_loss": 0.008056640625,
|
| 59 |
+
"retain_loss": 1.6570091247558594e-05,
|
| 60 |
+
"step": 10
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.03755868544600939,
|
| 64 |
+
"forget_loss": 0.0081787109375,
|
| 65 |
+
"retain_loss": 2.2649765014648438e-05,
|
| 66 |
+
"step": 10
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.03755868544600939,
|
| 70 |
+
"forget_loss": 0.00823974609375,
|
| 71 |
+
"retain_loss": 2.1457672119140625e-05,
|
| 72 |
+
"step": 10
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07511737089201878,
|
| 76 |
+
"grad_norm": 0.047607421875,
|
| 77 |
+
"learning_rate": 9.5e-06,
|
| 78 |
+
"loss": 0.00635223388671875,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07511737089201878,
|
| 83 |
+
"forget_loss": 0.003662109375,
|
| 84 |
+
"retain_loss": 0.00022602081298828125,
|
| 85 |
+
"step": 20
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.07511737089201878,
|
| 89 |
+
"forget_loss": 0.004425048828125,
|
| 90 |
+
"retain_loss": 0.0002880096435546875,
|
| 91 |
+
"step": 20
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.07511737089201878,
|
| 95 |
+
"forget_loss": 0.0037078857421875,
|
| 96 |
+
"retain_loss": 0.00020885467529296875,
|
| 97 |
+
"step": 20
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.07511737089201878,
|
| 101 |
+
"forget_loss": 0.00421142578125,
|
| 102 |
+
"retain_loss": 0.00021457672119140625,
|
| 103 |
+
"step": 20
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.11267605633802817,
|
| 107 |
+
"grad_norm": 0.123046875,
|
| 108 |
+
"learning_rate": 9.82490272373541e-06,
|
| 109 |
+
"loss": 0.00320587158203125,
|
| 110 |
+
"step": 30
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.11267605633802817,
|
| 114 |
+
"forget_loss": 0.0018310546875,
|
| 115 |
+
"retain_loss": 0.00054168701171875,
|
| 116 |
+
"step": 30
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.11267605633802817,
|
| 120 |
+
"forget_loss": 0.00177001953125,
|
| 121 |
+
"retain_loss": 0.0004177093505859375,
|
| 122 |
+
"step": 30
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.11267605633802817,
|
| 126 |
+
"forget_loss": 0.00177001953125,
|
| 127 |
+
"retain_loss": 0.0002956390380859375,
|
| 128 |
+
"step": 30
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.11267605633802817,
|
| 132 |
+
"forget_loss": 0.0016937255859375,
|
| 133 |
+
"retain_loss": 0.000331878662109375,
|
| 134 |
+
"step": 30
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.15023474178403756,
|
| 138 |
+
"grad_norm": 0.05517578125,
|
| 139 |
+
"learning_rate": 9.63035019455253e-06,
|
| 140 |
+
"loss": 0.0016145706176757812,
|
| 141 |
+
"step": 40
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.15023474178403756,
|
| 145 |
+
"forget_loss": 0.0007781982421875,
|
| 146 |
+
"retain_loss": 0.00021648406982421875,
|
| 147 |
+
"step": 40
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.15023474178403756,
|
| 151 |
+
"forget_loss": 0.000762939453125,
|
| 152 |
+
"retain_loss": 0.0004100799560546875,
|
| 153 |
+
"step": 40
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.15023474178403756,
|
| 157 |
+
"forget_loss": 0.000766754150390625,
|
| 158 |
+
"retain_loss": 0.000331878662109375,
|
| 159 |
+
"step": 40
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.15023474178403756,
|
| 163 |
+
"forget_loss": 0.0007781982421875,
|
| 164 |
+
"retain_loss": 0.00038909912109375,
|
| 165 |
+
"step": 40
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.18779342723004694,
|
| 169 |
+
"grad_norm": 0.1416015625,
|
| 170 |
+
"learning_rate": 9.435797665369651e-06,
|
| 171 |
+
"loss": 0.000954437255859375,
|
| 172 |
+
"step": 50
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.18779342723004694,
|
| 176 |
+
"forget_loss": 0.0003452301025390625,
|
| 177 |
+
"retain_loss": 0.00020599365234375,
|
| 178 |
+
"step": 50
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.18779342723004694,
|
| 182 |
+
"forget_loss": 0.0003814697265625,
|
| 183 |
+
"retain_loss": 0.00020694732666015625,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.18779342723004694,
|
| 188 |
+
"forget_loss": 0.000415802001953125,
|
| 189 |
+
"retain_loss": 0.0003604888916015625,
|
| 190 |
+
"step": 50
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18779342723004694,
|
| 194 |
+
"forget_loss": 0.0003719329833984375,
|
| 195 |
+
"retain_loss": 0.0001964569091796875,
|
| 196 |
+
"step": 50
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.22535211267605634,
|
| 200 |
+
"grad_norm": 0.06396484375,
|
| 201 |
+
"learning_rate": 9.241245136186772e-06,
|
| 202 |
+
"loss": 0.0008329868316650391,
|
| 203 |
+
"step": 60
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.22535211267605634,
|
| 207 |
+
"forget_loss": 0.0002574920654296875,
|
| 208 |
+
"retain_loss": 0.0002727508544921875,
|
| 209 |
+
"step": 60
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.22535211267605634,
|
| 213 |
+
"forget_loss": 0.00019931793212890625,
|
| 214 |
+
"retain_loss": 0.00010919570922851562,
|
| 215 |
+
"step": 60
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.22535211267605634,
|
| 219 |
+
"forget_loss": 0.0001983642578125,
|
| 220 |
+
"retain_loss": 0.000202178955078125,
|
| 221 |
+
"step": 60
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.22535211267605634,
|
| 225 |
+
"forget_loss": 0.00023365020751953125,
|
| 226 |
+
"retain_loss": 0.0002002716064453125,
|
| 227 |
+
"step": 60
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.26291079812206575,
|
| 231 |
+
"grad_norm": 0.28515625,
|
| 232 |
+
"learning_rate": 9.046692607003892e-06,
|
| 233 |
+
"loss": 0.0006081581115722657,
|
| 234 |
+
"step": 70
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.26291079812206575,
|
| 238 |
+
"forget_loss": 0.00018405914306640625,
|
| 239 |
+
"retain_loss": 0.0002689361572265625,
|
| 240 |
+
"step": 70
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.26291079812206575,
|
| 244 |
+
"forget_loss": 0.0001811981201171875,
|
| 245 |
+
"retain_loss": 8.535385131835938e-05,
|
| 246 |
+
"step": 70
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.26291079812206575,
|
| 250 |
+
"forget_loss": 0.00022792816162109375,
|
| 251 |
+
"retain_loss": 0.00010728836059570312,
|
| 252 |
+
"step": 70
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.26291079812206575,
|
| 256 |
+
"forget_loss": 0.0003643035888671875,
|
| 257 |
+
"retain_loss": 0.00011777877807617188,
|
| 258 |
+
"step": 70
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.3004694835680751,
|
| 262 |
+
"grad_norm": 0.07177734375,
|
| 263 |
+
"learning_rate": 8.852140077821012e-06,
|
| 264 |
+
"loss": 0.0004550933837890625,
|
| 265 |
+
"step": 80
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.3004694835680751,
|
| 269 |
+
"forget_loss": 0.0002155303955078125,
|
| 270 |
+
"retain_loss": 0.000110626220703125,
|
| 271 |
+
"step": 80
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.3004694835680751,
|
| 275 |
+
"forget_loss": 0.00014019012451171875,
|
| 276 |
+
"retain_loss": 6.723403930664062e-05,
|
| 277 |
+
"step": 80
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.3004694835680751,
|
| 281 |
+
"forget_loss": 0.0001773834228515625,
|
| 282 |
+
"retain_loss": 0.000331878662109375,
|
| 283 |
+
"step": 80
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3004694835680751,
|
| 287 |
+
"forget_loss": 0.00016498565673828125,
|
| 288 |
+
"retain_loss": 9.489059448242188e-05,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3380281690140845,
|
| 293 |
+
"grad_norm": 0.0810546875,
|
| 294 |
+
"learning_rate": 8.657587548638134e-06,
|
| 295 |
+
"loss": 0.0003980875015258789,
|
| 296 |
+
"step": 90
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3380281690140845,
|
| 300 |
+
"forget_loss": 0.00015735626220703125,
|
| 301 |
+
"retain_loss": 0.0001544952392578125,
|
| 302 |
+
"step": 90
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.3380281690140845,
|
| 306 |
+
"forget_loss": 0.000133514404296875,
|
| 307 |
+
"retain_loss": 0.00080108642578125,
|
| 308 |
+
"step": 90
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.3380281690140845,
|
| 312 |
+
"forget_loss": 0.00014781951904296875,
|
| 313 |
+
"retain_loss": 0.000431060791015625,
|
| 314 |
+
"step": 90
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.3380281690140845,
|
| 318 |
+
"forget_loss": 0.00014400482177734375,
|
| 319 |
+
"retain_loss": 0.00031280517578125,
|
| 320 |
+
"step": 90
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.3755868544600939,
|
| 324 |
+
"grad_norm": 0.025146484375,
|
| 325 |
+
"learning_rate": 8.463035019455253e-06,
|
| 326 |
+
"loss": 0.00032124519348144533,
|
| 327 |
+
"step": 100
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.3755868544600939,
|
| 331 |
+
"forget_loss": 0.00015163421630859375,
|
| 332 |
+
"retain_loss": 0.00015926361083984375,
|
| 333 |
+
"step": 100
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.3755868544600939,
|
| 337 |
+
"forget_loss": 0.0001506805419921875,
|
| 338 |
+
"retain_loss": 0.00011587142944335938,
|
| 339 |
+
"step": 100
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3755868544600939,
|
| 343 |
+
"forget_loss": 0.00013065338134765625,
|
| 344 |
+
"retain_loss": 0.000125885009765625,
|
| 345 |
+
"step": 100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3755868544600939,
|
| 349 |
+
"forget_loss": 0.0001239776611328125,
|
| 350 |
+
"retain_loss": 0.0001277923583984375,
|
| 351 |
+
"step": 100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.4131455399061033,
|
| 355 |
+
"grad_norm": 0.06884765625,
|
| 356 |
+
"learning_rate": 8.268482490272373e-06,
|
| 357 |
+
"loss": 0.00034072399139404295,
|
| 358 |
+
"step": 110
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.4131455399061033,
|
| 362 |
+
"forget_loss": 0.0001544952392578125,
|
| 363 |
+
"retain_loss": 7.2479248046875e-05,
|
| 364 |
+
"step": 110
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.4131455399061033,
|
| 368 |
+
"forget_loss": 0.0001373291015625,
|
| 369 |
+
"retain_loss": 0.000270843505859375,
|
| 370 |
+
"step": 110
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.4131455399061033,
|
| 374 |
+
"forget_loss": 0.0001125335693359375,
|
| 375 |
+
"retain_loss": 9.965896606445312e-05,
|
| 376 |
+
"step": 110
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.4131455399061033,
|
| 380 |
+
"forget_loss": 0.00011205673217773438,
|
| 381 |
+
"retain_loss": 6.866455078125e-05,
|
| 382 |
+
"step": 110
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.4507042253521127,
|
| 386 |
+
"grad_norm": 0.08544921875,
|
| 387 |
+
"learning_rate": 8.073929961089495e-06,
|
| 388 |
+
"loss": 0.0002372264862060547,
|
| 389 |
+
"step": 120
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.4507042253521127,
|
| 393 |
+
"forget_loss": 0.000179290771484375,
|
| 394 |
+
"retain_loss": 7.05718994140625e-05,
|
| 395 |
+
"step": 120
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.4507042253521127,
|
| 399 |
+
"forget_loss": 0.0001277923583984375,
|
| 400 |
+
"retain_loss": 9.250640869140625e-05,
|
| 401 |
+
"step": 120
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4507042253521127,
|
| 405 |
+
"forget_loss": 0.00013446807861328125,
|
| 406 |
+
"retain_loss": 8.630752563476562e-05,
|
| 407 |
+
"step": 120
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.4507042253521127,
|
| 411 |
+
"forget_loss": 0.0001735687255859375,
|
| 412 |
+
"retain_loss": 8.106231689453125e-05,
|
| 413 |
+
"step": 120
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.48826291079812206,
|
| 417 |
+
"grad_norm": 0.0361328125,
|
| 418 |
+
"learning_rate": 7.879377431906614e-06,
|
| 419 |
+
"loss": 0.0002105236053466797,
|
| 420 |
+
"step": 130
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.48826291079812206,
|
| 424 |
+
"forget_loss": 0.00010776519775390625,
|
| 425 |
+
"retain_loss": 5.221366882324219e-05,
|
| 426 |
+
"step": 130
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.48826291079812206,
|
| 430 |
+
"forget_loss": 0.00010728836059570312,
|
| 431 |
+
"retain_loss": 5.1021575927734375e-05,
|
| 432 |
+
"step": 130
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.48826291079812206,
|
| 436 |
+
"forget_loss": 0.00010204315185546875,
|
| 437 |
+
"retain_loss": 0.00011110305786132812,
|
| 438 |
+
"step": 130
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.48826291079812206,
|
| 442 |
+
"forget_loss": 0.000156402587890625,
|
| 443 |
+
"retain_loss": 7.05718994140625e-05,
|
| 444 |
+
"step": 130
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5258215962441315,
|
| 448 |
+
"grad_norm": 0.035400390625,
|
| 449 |
+
"learning_rate": 7.684824902723736e-06,
|
| 450 |
+
"loss": 0.000272369384765625,
|
| 451 |
+
"step": 140
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5258215962441315,
|
| 455 |
+
"forget_loss": 0.0001087188720703125,
|
| 456 |
+
"retain_loss": 7.009506225585938e-05,
|
| 457 |
+
"step": 140
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5258215962441315,
|
| 461 |
+
"forget_loss": 0.0001239776611328125,
|
| 462 |
+
"retain_loss": 7.963180541992188e-05,
|
| 463 |
+
"step": 140
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.5258215962441315,
|
| 467 |
+
"forget_loss": 0.00017070770263671875,
|
| 468 |
+
"retain_loss": 6.29425048828125e-05,
|
| 469 |
+
"step": 140
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.5258215962441315,
|
| 473 |
+
"forget_loss": 0.00013446807861328125,
|
| 474 |
+
"retain_loss": 5.5789947509765625e-05,
|
| 475 |
+
"step": 140
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.5633802816901409,
|
| 479 |
+
"grad_norm": 0.0177001953125,
|
| 480 |
+
"learning_rate": 7.4902723735408565e-06,
|
| 481 |
+
"loss": 0.00019230842590332032,
|
| 482 |
+
"step": 150
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.5633802816901409,
|
| 486 |
+
"forget_loss": 0.00010585784912109375,
|
| 487 |
+
"retain_loss": 5.793571472167969e-05,
|
| 488 |
+
"step": 150
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.5633802816901409,
|
| 492 |
+
"forget_loss": 0.00011157989501953125,
|
| 493 |
+
"retain_loss": 0.00015163421630859375,
|
| 494 |
+
"step": 150
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.5633802816901409,
|
| 498 |
+
"forget_loss": 8.821487426757812e-05,
|
| 499 |
+
"retain_loss": 6.866455078125e-05,
|
| 500 |
+
"step": 150
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5633802816901409,
|
| 504 |
+
"forget_loss": 9.059906005859375e-05,
|
| 505 |
+
"retain_loss": 0.00010251998901367188,
|
| 506 |
+
"step": 150
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.6009389671361502,
|
| 510 |
+
"grad_norm": 0.06591796875,
|
| 511 |
+
"learning_rate": 7.295719844357977e-06,
|
| 512 |
+
"loss": 0.00020923614501953126,
|
| 513 |
+
"step": 160
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.6009389671361502,
|
| 517 |
+
"forget_loss": 0.00010156631469726562,
|
| 518 |
+
"retain_loss": 8.535385131835938e-05,
|
| 519 |
+
"step": 160
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.6009389671361502,
|
| 523 |
+
"forget_loss": 9.441375732421875e-05,
|
| 524 |
+
"retain_loss": 7.05718994140625e-05,
|
| 525 |
+
"step": 160
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.6009389671361502,
|
| 529 |
+
"forget_loss": 9.489059448242188e-05,
|
| 530 |
+
"retain_loss": 6.151199340820312e-05,
|
| 531 |
+
"step": 160
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.6009389671361502,
|
| 535 |
+
"forget_loss": 9.679794311523438e-05,
|
| 536 |
+
"retain_loss": 5.984306335449219e-05,
|
| 537 |
+
"step": 160
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.6384976525821596,
|
| 541 |
+
"grad_norm": 0.0152587890625,
|
| 542 |
+
"learning_rate": 7.101167315175098e-06,
|
| 543 |
+
"loss": 0.0001931428909301758,
|
| 544 |
+
"step": 170
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.6384976525821596,
|
| 548 |
+
"forget_loss": 9.489059448242188e-05,
|
| 549 |
+
"retain_loss": 6.079673767089844e-05,
|
| 550 |
+
"step": 170
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.6384976525821596,
|
| 554 |
+
"forget_loss": 0.000110626220703125,
|
| 555 |
+
"retain_loss": 7.104873657226562e-05,
|
| 556 |
+
"step": 170
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6384976525821596,
|
| 560 |
+
"forget_loss": 0.00010776519775390625,
|
| 561 |
+
"retain_loss": 0.00011539459228515625,
|
| 562 |
+
"step": 170
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6384976525821596,
|
| 566 |
+
"forget_loss": 9.441375732421875e-05,
|
| 567 |
+
"retain_loss": 0.00011682510375976562,
|
| 568 |
+
"step": 170
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.676056338028169,
|
| 572 |
+
"grad_norm": 0.0306396484375,
|
| 573 |
+
"learning_rate": 6.9066147859922184e-06,
|
| 574 |
+
"loss": 0.00021104812622070314,
|
| 575 |
+
"step": 180
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.676056338028169,
|
| 579 |
+
"forget_loss": 0.000125885009765625,
|
| 580 |
+
"retain_loss": 0.0001239776611328125,
|
| 581 |
+
"step": 180
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.676056338028169,
|
| 585 |
+
"forget_loss": 0.00010395050048828125,
|
| 586 |
+
"retain_loss": 8.96453857421875e-05,
|
| 587 |
+
"step": 180
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 0.676056338028169,
|
| 591 |
+
"forget_loss": 8.869171142578125e-05,
|
| 592 |
+
"retain_loss": 8.20159912109375e-05,
|
| 593 |
+
"step": 180
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.676056338028169,
|
| 597 |
+
"forget_loss": 9.107589721679688e-05,
|
| 598 |
+
"retain_loss": 5.316734313964844e-05,
|
| 599 |
+
"step": 180
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.7136150234741784,
|
| 603 |
+
"grad_norm": 0.0230712890625,
|
| 604 |
+
"learning_rate": 6.712062256809338e-06,
|
| 605 |
+
"loss": 0.0001739978790283203,
|
| 606 |
+
"step": 190
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.7136150234741784,
|
| 610 |
+
"forget_loss": 9.059906005859375e-05,
|
| 611 |
+
"retain_loss": 5.793571472167969e-05,
|
| 612 |
+
"step": 190
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7136150234741784,
|
| 616 |
+
"forget_loss": 9.5367431640625e-05,
|
| 617 |
+
"retain_loss": 5.0067901611328125e-05,
|
| 618 |
+
"step": 190
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7136150234741784,
|
| 622 |
+
"forget_loss": 8.869171142578125e-05,
|
| 623 |
+
"retain_loss": 5.7220458984375e-05,
|
| 624 |
+
"step": 190
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.7136150234741784,
|
| 628 |
+
"forget_loss": 0.00010919570922851562,
|
| 629 |
+
"retain_loss": 4.9591064453125e-05,
|
| 630 |
+
"step": 190
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.7511737089201878,
|
| 634 |
+
"grad_norm": 0.01068115234375,
|
| 635 |
+
"learning_rate": 6.51750972762646e-06,
|
| 636 |
+
"loss": 0.00019364356994628907,
|
| 637 |
+
"step": 200
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 0.7511737089201878,
|
| 641 |
+
"forget_loss": 0.00010204315185546875,
|
| 642 |
+
"retain_loss": 5.14984130859375e-05,
|
| 643 |
+
"step": 200
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.7511737089201878,
|
| 647 |
+
"forget_loss": 8.249282836914062e-05,
|
| 648 |
+
"retain_loss": 4.887580871582031e-05,
|
| 649 |
+
"step": 200
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.7511737089201878,
|
| 653 |
+
"forget_loss": 7.772445678710938e-05,
|
| 654 |
+
"retain_loss": 4.076957702636719e-05,
|
| 655 |
+
"step": 200
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.7511737089201878,
|
| 659 |
+
"forget_loss": 0.00012683868408203125,
|
| 660 |
+
"retain_loss": 6.341934204101562e-05,
|
| 661 |
+
"step": 200
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.7887323943661971,
|
| 665 |
+
"grad_norm": 0.016357421875,
|
| 666 |
+
"learning_rate": 6.32295719844358e-06,
|
| 667 |
+
"loss": 0.00015709400177001953,
|
| 668 |
+
"step": 210
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.7887323943661971,
|
| 672 |
+
"forget_loss": 0.00010585784912109375,
|
| 673 |
+
"retain_loss": 7.009506225585938e-05,
|
| 674 |
+
"step": 210
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7887323943661971,
|
| 678 |
+
"forget_loss": 8.58306884765625e-05,
|
| 679 |
+
"retain_loss": 4.2438507080078125e-05,
|
| 680 |
+
"step": 210
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.7887323943661971,
|
| 684 |
+
"forget_loss": 8.440017700195312e-05,
|
| 685 |
+
"retain_loss": 4.744529724121094e-05,
|
| 686 |
+
"step": 210
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.7887323943661971,
|
| 690 |
+
"forget_loss": 0.0001010894775390625,
|
| 691 |
+
"retain_loss": 4.4345855712890625e-05,
|
| 692 |
+
"step": 210
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.8262910798122066,
|
| 696 |
+
"grad_norm": 0.01806640625,
|
| 697 |
+
"learning_rate": 6.128404669260701e-06,
|
| 698 |
+
"loss": 0.0001538991928100586,
|
| 699 |
+
"step": 220
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.8262910798122066,
|
| 703 |
+
"forget_loss": 9.72747802734375e-05,
|
| 704 |
+
"retain_loss": 4.8160552978515625e-05,
|
| 705 |
+
"step": 220
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.8262910798122066,
|
| 709 |
+
"forget_loss": 8.0108642578125e-05,
|
| 710 |
+
"retain_loss": 7.200241088867188e-05,
|
| 711 |
+
"step": 220
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.8262910798122066,
|
| 715 |
+
"forget_loss": 7.82012939453125e-05,
|
| 716 |
+
"retain_loss": 4.3392181396484375e-05,
|
| 717 |
+
"step": 220
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.8262910798122066,
|
| 721 |
+
"forget_loss": 0.0001049041748046875,
|
| 722 |
+
"retain_loss": 7.009506225585938e-05,
|
| 723 |
+
"step": 220
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.863849765258216,
|
| 727 |
+
"grad_norm": 0.0184326171875,
|
| 728 |
+
"learning_rate": 5.933852140077822e-06,
|
| 729 |
+
"loss": 0.00015753507614135742,
|
| 730 |
+
"step": 230
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.863849765258216,
|
| 734 |
+
"forget_loss": 8.678436279296875e-05,
|
| 735 |
+
"retain_loss": 4.38690185546875e-05,
|
| 736 |
+
"step": 230
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.863849765258216,
|
| 740 |
+
"forget_loss": 8.344650268554688e-05,
|
| 741 |
+
"retain_loss": 0.000942230224609375,
|
| 742 |
+
"step": 230
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.863849765258216,
|
| 746 |
+
"forget_loss": 9.1552734375e-05,
|
| 747 |
+
"retain_loss": 4.410743713378906e-05,
|
| 748 |
+
"step": 230
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 0.863849765258216,
|
| 752 |
+
"forget_loss": 8.296966552734375e-05,
|
| 753 |
+
"retain_loss": 7.486343383789062e-05,
|
| 754 |
+
"step": 230
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.9014084507042254,
|
| 758 |
+
"grad_norm": 0.0238037109375,
|
| 759 |
+
"learning_rate": 5.7392996108949415e-06,
|
| 760 |
+
"loss": 0.00017303228378295898,
|
| 761 |
+
"step": 240
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.9014084507042254,
|
| 765 |
+
"forget_loss": 8.296966552734375e-05,
|
| 766 |
+
"retain_loss": 4.291534423828125e-05,
|
| 767 |
+
"step": 240
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.9014084507042254,
|
| 771 |
+
"forget_loss": 8.96453857421875e-05,
|
| 772 |
+
"retain_loss": 3.814697265625e-05,
|
| 773 |
+
"step": 240
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.9014084507042254,
|
| 777 |
+
"forget_loss": 8.726119995117188e-05,
|
| 778 |
+
"retain_loss": 4.839897155761719e-05,
|
| 779 |
+
"step": 240
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9014084507042254,
|
| 783 |
+
"forget_loss": 7.200241088867188e-05,
|
| 784 |
+
"retain_loss": 4.839897155761719e-05,
|
| 785 |
+
"step": 240
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.9389671361502347,
|
| 789 |
+
"grad_norm": 0.05908203125,
|
| 790 |
+
"learning_rate": 5.544747081712063e-06,
|
| 791 |
+
"loss": 0.00014340877532958984,
|
| 792 |
+
"step": 250
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.9389671361502347,
|
| 796 |
+
"forget_loss": 0.00010442733764648438,
|
| 797 |
+
"retain_loss": 5.626678466796875e-05,
|
| 798 |
+
"step": 250
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.9389671361502347,
|
| 802 |
+
"forget_loss": 8.726119995117188e-05,
|
| 803 |
+
"retain_loss": 9.059906005859375e-05,
|
| 804 |
+
"step": 250
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.9389671361502347,
|
| 808 |
+
"forget_loss": 8.344650268554688e-05,
|
| 809 |
+
"retain_loss": 3.647804260253906e-05,
|
| 810 |
+
"step": 250
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.9389671361502347,
|
| 814 |
+
"forget_loss": 8.821487426757812e-05,
|
| 815 |
+
"retain_loss": 4.220008850097656e-05,
|
| 816 |
+
"step": 250
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.9765258215962441,
|
| 820 |
+
"grad_norm": 0.0181884765625,
|
| 821 |
+
"learning_rate": 5.350194552529184e-06,
|
| 822 |
+
"loss": 0.00015048980712890626,
|
| 823 |
+
"step": 260
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.9765258215962441,
|
| 827 |
+
"forget_loss": 8.153915405273438e-05,
|
| 828 |
+
"retain_loss": 6.580352783203125e-05,
|
| 829 |
+
"step": 260
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9765258215962441,
|
| 833 |
+
"forget_loss": 8.7738037109375e-05,
|
| 834 |
+
"retain_loss": 3.6716461181640625e-05,
|
| 835 |
+
"step": 260
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.9765258215962441,
|
| 839 |
+
"forget_loss": 9.918212890625e-05,
|
| 840 |
+
"retain_loss": 4.363059997558594e-05,
|
| 841 |
+
"step": 260
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.9765258215962441,
|
| 845 |
+
"forget_loss": 9.870529174804688e-05,
|
| 846 |
+
"retain_loss": 0.0001506805419921875,
|
| 847 |
+
"step": 260
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 1.0112676056338028,
|
| 851 |
+
"grad_norm": 0.00848388671875,
|
| 852 |
+
"learning_rate": 5.155642023346303e-06,
|
| 853 |
+
"loss": 0.00013761520385742189,
|
| 854 |
+
"step": 270
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.0112676056338028,
|
| 858 |
+
"forget_loss": 6.866455078125e-05,
|
| 859 |
+
"retain_loss": 5.9604644775390625e-05,
|
| 860 |
+
"step": 270
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.0112676056338028,
|
| 864 |
+
"forget_loss": 8.249282836914062e-05,
|
| 865 |
+
"retain_loss": 3.5762786865234375e-05,
|
| 866 |
+
"step": 270
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.0112676056338028,
|
| 870 |
+
"forget_loss": 8.392333984375e-05,
|
| 871 |
+
"retain_loss": 6.246566772460938e-05,
|
| 872 |
+
"step": 270
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.0112676056338028,
|
| 876 |
+
"forget_loss": 8.916854858398438e-05,
|
| 877 |
+
"retain_loss": 3.695487976074219e-05,
|
| 878 |
+
"step": 270
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.0488262910798123,
|
| 882 |
+
"grad_norm": 0.0137939453125,
|
| 883 |
+
"learning_rate": 4.961089494163425e-06,
|
| 884 |
+
"loss": 0.00013872385025024415,
|
| 885 |
+
"step": 280
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.0488262910798123,
|
| 889 |
+
"forget_loss": 7.772445678710938e-05,
|
| 890 |
+
"retain_loss": 4.982948303222656e-05,
|
| 891 |
+
"step": 280
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.0488262910798123,
|
| 895 |
+
"forget_loss": 6.771087646484375e-05,
|
| 896 |
+
"retain_loss": 4.5299530029296875e-05,
|
| 897 |
+
"step": 280
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 1.0488262910798123,
|
| 901 |
+
"forget_loss": 6.961822509765625e-05,
|
| 902 |
+
"retain_loss": 5.14984130859375e-05,
|
| 903 |
+
"step": 280
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 1.0488262910798123,
|
| 907 |
+
"forget_loss": 7.05718994140625e-05,
|
| 908 |
+
"retain_loss": 5.245208740234375e-05,
|
| 909 |
+
"step": 280
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 1.0863849765258216,
|
| 913 |
+
"grad_norm": 0.01275634765625,
|
| 914 |
+
"learning_rate": 4.7665369649805455e-06,
|
| 915 |
+
"loss": 0.00014275312423706055,
|
| 916 |
+
"step": 290
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 1.0863849765258216,
|
| 920 |
+
"forget_loss": 8.20159912109375e-05,
|
| 921 |
+
"retain_loss": 4.38690185546875e-05,
|
| 922 |
+
"step": 290
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.0863849765258216,
|
| 926 |
+
"forget_loss": 7.62939453125e-05,
|
| 927 |
+
"retain_loss": 5.745887756347656e-05,
|
| 928 |
+
"step": 290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.0863849765258216,
|
| 932 |
+
"forget_loss": 9.059906005859375e-05,
|
| 933 |
+
"retain_loss": 5.269050598144531e-05,
|
| 934 |
+
"step": 290
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.0863849765258216,
|
| 938 |
+
"forget_loss": 8.106231689453125e-05,
|
| 939 |
+
"retain_loss": 5.507469177246094e-05,
|
| 940 |
+
"step": 290
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.123943661971831,
|
| 944 |
+
"grad_norm": 0.0140380859375,
|
| 945 |
+
"learning_rate": 4.571984435797665e-06,
|
| 946 |
+
"loss": 0.00013855695724487304,
|
| 947 |
+
"step": 300
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.123943661971831,
|
| 951 |
+
"forget_loss": 8.344650268554688e-05,
|
| 952 |
+
"retain_loss": 3.695487976074219e-05,
|
| 953 |
+
"step": 300
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 1.123943661971831,
|
| 957 |
+
"forget_loss": 8.249282836914062e-05,
|
| 958 |
+
"retain_loss": 4.100799560546875e-05,
|
| 959 |
+
"step": 300
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 1.123943661971831,
|
| 963 |
+
"forget_loss": 0.00010395050048828125,
|
| 964 |
+
"retain_loss": 3.504753112792969e-05,
|
| 965 |
+
"step": 300
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.123943661971831,
|
| 969 |
+
"forget_loss": 6.4849853515625e-05,
|
| 970 |
+
"retain_loss": 5.5789947509765625e-05,
|
| 971 |
+
"step": 300
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.1615023474178403,
|
| 975 |
+
"grad_norm": 0.019775390625,
|
| 976 |
+
"learning_rate": 4.377431906614786e-06,
|
| 977 |
+
"loss": 0.0001352071762084961,
|
| 978 |
+
"step": 310
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.1615023474178403,
|
| 982 |
+
"forget_loss": 7.62939453125e-05,
|
| 983 |
+
"retain_loss": 7.772445678710938e-05,
|
| 984 |
+
"step": 310
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.1615023474178403,
|
| 988 |
+
"forget_loss": 8.58306884765625e-05,
|
| 989 |
+
"retain_loss": 5.459785461425781e-05,
|
| 990 |
+
"step": 310
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1615023474178403,
|
| 994 |
+
"forget_loss": 7.05718994140625e-05,
|
| 995 |
+
"retain_loss": 7.05718994140625e-05,
|
| 996 |
+
"step": 310
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.1615023474178403,
|
| 1000 |
+
"forget_loss": 7.82012939453125e-05,
|
| 1001 |
+
"retain_loss": 3.337860107421875e-05,
|
| 1002 |
+
"step": 310
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 1.1990610328638498,
|
| 1006 |
+
"grad_norm": 0.020751953125,
|
| 1007 |
+
"learning_rate": 4.1828793774319074e-06,
|
| 1008 |
+
"loss": 0.000127410888671875,
|
| 1009 |
+
"step": 320
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 1.1990610328638498,
|
| 1013 |
+
"forget_loss": 7.867813110351562e-05,
|
| 1014 |
+
"retain_loss": 3.695487976074219e-05,
|
| 1015 |
+
"step": 320
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 1.1990610328638498,
|
| 1019 |
+
"forget_loss": 8.249282836914062e-05,
|
| 1020 |
+
"retain_loss": 5.412101745605469e-05,
|
| 1021 |
+
"step": 320
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 1.1990610328638498,
|
| 1025 |
+
"forget_loss": 9.775161743164062e-05,
|
| 1026 |
+
"retain_loss": 5.340576171875e-05,
|
| 1027 |
+
"step": 320
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 1.1990610328638498,
|
| 1031 |
+
"forget_loss": 7.152557373046875e-05,
|
| 1032 |
+
"retain_loss": 8.106231689453125e-05,
|
| 1033 |
+
"step": 320
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.236619718309859,
|
| 1037 |
+
"grad_norm": 0.0091552734375,
|
| 1038 |
+
"learning_rate": 3.988326848249028e-06,
|
| 1039 |
+
"loss": 0.00014822483062744142,
|
| 1040 |
+
"step": 330
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.236619718309859,
|
| 1044 |
+
"forget_loss": 9.72747802734375e-05,
|
| 1045 |
+
"retain_loss": 4.506111145019531e-05,
|
| 1046 |
+
"step": 330
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.236619718309859,
|
| 1050 |
+
"forget_loss": 0.00010919570922851562,
|
| 1051 |
+
"retain_loss": 3.2901763916015625e-05,
|
| 1052 |
+
"step": 330
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.236619718309859,
|
| 1056 |
+
"forget_loss": 6.723403930664062e-05,
|
| 1057 |
+
"retain_loss": 4.029273986816406e-05,
|
| 1058 |
+
"step": 330
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1.236619718309859,
|
| 1062 |
+
"forget_loss": 6.914138793945312e-05,
|
| 1063 |
+
"retain_loss": 7.724761962890625e-05,
|
| 1064 |
+
"step": 330
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 1.2741784037558685,
|
| 1068 |
+
"grad_norm": 0.011962890625,
|
| 1069 |
+
"learning_rate": 3.793774319066148e-06,
|
| 1070 |
+
"loss": 0.0001390695571899414,
|
| 1071 |
+
"step": 340
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 1.2741784037558685,
|
| 1075 |
+
"forget_loss": 8.535385131835938e-05,
|
| 1076 |
+
"retain_loss": 4.506111145019531e-05,
|
| 1077 |
+
"step": 340
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.2741784037558685,
|
| 1081 |
+
"forget_loss": 6.866455078125e-05,
|
| 1082 |
+
"retain_loss": 3.2901763916015625e-05,
|
| 1083 |
+
"step": 340
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.2741784037558685,
|
| 1087 |
+
"forget_loss": 6.723403930664062e-05,
|
| 1088 |
+
"retain_loss": 5.0067901611328125e-05,
|
| 1089 |
+
"step": 340
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.2741784037558685,
|
| 1093 |
+
"forget_loss": 8.344650268554688e-05,
|
| 1094 |
+
"retain_loss": 3.5762786865234375e-05,
|
| 1095 |
+
"step": 340
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.3117370892018778,
|
| 1099 |
+
"grad_norm": 0.01239013671875,
|
| 1100 |
+
"learning_rate": 3.599221789883269e-06,
|
| 1101 |
+
"loss": 0.00013332366943359376,
|
| 1102 |
+
"step": 350
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.3117370892018778,
|
| 1106 |
+
"forget_loss": 7.772445678710938e-05,
|
| 1107 |
+
"retain_loss": 3.0040740966796875e-05,
|
| 1108 |
+
"step": 350
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.3117370892018778,
|
| 1112 |
+
"forget_loss": 7.963180541992188e-05,
|
| 1113 |
+
"retain_loss": 4.363059997558594e-05,
|
| 1114 |
+
"step": 350
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 1.3117370892018778,
|
| 1118 |
+
"forget_loss": 8.630752563476562e-05,
|
| 1119 |
+
"retain_loss": 4.6253204345703125e-05,
|
| 1120 |
+
"step": 350
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 1.3117370892018778,
|
| 1124 |
+
"forget_loss": 8.487701416015625e-05,
|
| 1125 |
+
"retain_loss": 5.030632019042969e-05,
|
| 1126 |
+
"step": 350
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 1.3492957746478873,
|
| 1130 |
+
"grad_norm": 0.01177978515625,
|
| 1131 |
+
"learning_rate": 3.4046692607003896e-06,
|
| 1132 |
+
"loss": 0.00012853145599365235,
|
| 1133 |
+
"step": 360
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.3492957746478873,
|
| 1137 |
+
"forget_loss": 0.00010395050048828125,
|
| 1138 |
+
"retain_loss": 3.3855438232421875e-05,
|
| 1139 |
+
"step": 360
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 1.3492957746478873,
|
| 1143 |
+
"forget_loss": 9.250640869140625e-05,
|
| 1144 |
+
"retain_loss": 4.553794860839844e-05,
|
| 1145 |
+
"step": 360
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.3492957746478873,
|
| 1149 |
+
"forget_loss": 7.009506225585938e-05,
|
| 1150 |
+
"retain_loss": 5.6743621826171875e-05,
|
| 1151 |
+
"step": 360
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.3492957746478873,
|
| 1155 |
+
"forget_loss": 6.961822509765625e-05,
|
| 1156 |
+
"retain_loss": 4.00543212890625e-05,
|
| 1157 |
+
"step": 360
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.3868544600938968,
|
| 1161 |
+
"grad_norm": 0.012451171875,
|
| 1162 |
+
"learning_rate": 3.2101167315175102e-06,
|
| 1163 |
+
"loss": 0.00012967586517333984,
|
| 1164 |
+
"step": 370
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.3868544600938968,
|
| 1168 |
+
"forget_loss": 8.296966552734375e-05,
|
| 1169 |
+
"retain_loss": 5.435943603515625e-05,
|
| 1170 |
+
"step": 370
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 1.3868544600938968,
|
| 1174 |
+
"forget_loss": 7.581710815429688e-05,
|
| 1175 |
+
"retain_loss": 3.647804260253906e-05,
|
| 1176 |
+
"step": 370
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 1.3868544600938968,
|
| 1180 |
+
"forget_loss": 7.05718994140625e-05,
|
| 1181 |
+
"retain_loss": 3.075599670410156e-05,
|
| 1182 |
+
"step": 370
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.3868544600938968,
|
| 1186 |
+
"forget_loss": 7.104873657226562e-05,
|
| 1187 |
+
"retain_loss": 0.00010776519775390625,
|
| 1188 |
+
"step": 370
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 1.424413145539906,
|
| 1192 |
+
"grad_norm": 0.0162353515625,
|
| 1193 |
+
"learning_rate": 3.0155642023346305e-06,
|
| 1194 |
+
"loss": 0.0001344442367553711,
|
| 1195 |
+
"step": 380
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 1.424413145539906,
|
| 1199 |
+
"forget_loss": 7.62939453125e-05,
|
| 1200 |
+
"retain_loss": 5.793571472167969e-05,
|
| 1201 |
+
"step": 380
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.424413145539906,
|
| 1205 |
+
"forget_loss": 6.246566772460938e-05,
|
| 1206 |
+
"retain_loss": 3.838539123535156e-05,
|
| 1207 |
+
"step": 380
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.424413145539906,
|
| 1211 |
+
"forget_loss": 0.00010538101196289062,
|
| 1212 |
+
"retain_loss": 4.291534423828125e-05,
|
| 1213 |
+
"step": 380
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.424413145539906,
|
| 1217 |
+
"forget_loss": 7.009506225585938e-05,
|
| 1218 |
+
"retain_loss": 4.482269287109375e-05,
|
| 1219 |
+
"step": 380
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 1.4619718309859155,
|
| 1223 |
+
"grad_norm": 0.00567626953125,
|
| 1224 |
+
"learning_rate": 2.821011673151751e-06,
|
| 1225 |
+
"loss": 0.00012153387069702148,
|
| 1226 |
+
"step": 390
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.4619718309859155,
|
| 1230 |
+
"forget_loss": 8.440017700195312e-05,
|
| 1231 |
+
"retain_loss": 5.7697296142578125e-05,
|
| 1232 |
+
"step": 390
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 1.4619718309859155,
|
| 1236 |
+
"forget_loss": 7.390975952148438e-05,
|
| 1237 |
+
"retain_loss": 5.793571472167969e-05,
|
| 1238 |
+
"step": 390
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.4619718309859155,
|
| 1242 |
+
"forget_loss": 9.489059448242188e-05,
|
| 1243 |
+
"retain_loss": 3.647804260253906e-05,
|
| 1244 |
+
"step": 390
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 1.4619718309859155,
|
| 1248 |
+
"forget_loss": 9.059906005859375e-05,
|
| 1249 |
+
"retain_loss": 3.504753112792969e-05,
|
| 1250 |
+
"step": 390
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.4995305164319248,
|
| 1254 |
+
"grad_norm": 0.0308837890625,
|
| 1255 |
+
"learning_rate": 2.6264591439688717e-06,
|
| 1256 |
+
"loss": 0.00013158321380615234,
|
| 1257 |
+
"step": 400
|
| 1258 |
+
}
|
| 1259 |
+
],
|
| 1260 |
+
"logging_steps": 10,
|
| 1261 |
+
"max_steps": 534,
|
| 1262 |
+
"num_input_tokens_seen": 0,
|
| 1263 |
+
"num_train_epochs": 2,
|
| 1264 |
+
"save_steps": 200,
|
| 1265 |
+
"stateful_callbacks": {
|
| 1266 |
+
"TrainerControl": {
|
| 1267 |
+
"args": {
|
| 1268 |
+
"should_epoch_stop": false,
|
| 1269 |
+
"should_evaluate": false,
|
| 1270 |
+
"should_log": false,
|
| 1271 |
+
"should_save": true,
|
| 1272 |
+
"should_training_stop": false
|
| 1273 |
+
},
|
| 1274 |
+
"attributes": {}
|
| 1275 |
+
}
|
| 1276 |
+
},
|
| 1277 |
+
"total_flos": 0.0,
|
| 1278 |
+
"train_batch_size": 8,
|
| 1279 |
+
"trial_name": null,
|
| 1280 |
+
"trial_params": null
|
| 1281 |
+
}
|
checkpoint-400/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cde3986b5d2902c3143ac3a5311b9443a600348ae4fbcaab8a20b0760c034ff1
|
| 3 |
+
size 5265
|
checkpoint-534/config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"LlamaForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 128000,
|
| 8 |
+
"dtype": "bfloat16",
|
| 9 |
+
"eos_token_id": 128001,
|
| 10 |
+
"head_dim": 128,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 4096,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 14336,
|
| 15 |
+
"max_position_embeddings": 131072,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 32,
|
| 19 |
+
"num_hidden_layers": 32,
|
| 20 |
+
"num_key_value_heads": 8,
|
| 21 |
+
"pad_token_id": 128001,
|
| 22 |
+
"pretraining_tp": 1,
|
| 23 |
+
"rms_norm_eps": 1e-05,
|
| 24 |
+
"rope_parameters": {
|
| 25 |
+
"factor": 8.0,
|
| 26 |
+
"high_freq_factor": 4.0,
|
| 27 |
+
"low_freq_factor": 1.0,
|
| 28 |
+
"original_max_position_embeddings": 8192,
|
| 29 |
+
"rope_theta": 500000.0,
|
| 30 |
+
"rope_type": "llama3"
|
| 31 |
+
},
|
| 32 |
+
"tie_word_embeddings": false,
|
| 33 |
+
"transformers_version": "5.6.0",
|
| 34 |
+
"use_cache": false,
|
| 35 |
+
"vocab_size": 128256
|
| 36 |
+
}
|
checkpoint-534/generation_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 128000,
|
| 4 |
+
"do_sample": true,
|
| 5 |
+
"eos_token_id": [
|
| 6 |
+
128001
|
| 7 |
+
],
|
| 8 |
+
"pad_token_id": 128001,
|
| 9 |
+
"temperature": 0.6,
|
| 10 |
+
"top_p": 0.9,
|
| 11 |
+
"transformers_version": "5.6.0"
|
| 12 |
+
}
|
checkpoint-534/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4b6fddb151d0b4bd2f8c5129e10e07f980a758209406b9308bd5b85a60af1f6
|
| 3 |
+
size 16060556616
|
checkpoint-534/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0fe69187eba836c52bbc34e344601eba624da43e0c24da7ee18dbb022f3e211d
|
| 3 |
+
size 9080995015
|
checkpoint-534/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c06081cf16221639a057723da86b40fbd9f838207c8739c97a81e7c17cffd882
|
| 3 |
+
size 14645
|
checkpoint-534/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dad762a3e6e973aadbcd756f264a89530c87974f025c1fa12a48c0e3e3dd8da5
|
| 3 |
+
size 1465
|
checkpoint-534/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
|
| 3 |
+
size 17209920
|
checkpoint-534/tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "<|begin_of_text|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "<|end_of_text|>",
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"local_files_only": false,
|
| 8 |
+
"model_input_names": [
|
| 9 |
+
"input_ids",
|
| 10 |
+
"attention_mask"
|
| 11 |
+
],
|
| 12 |
+
"model_max_length": 131072,
|
| 13 |
+
"pad_token": "<|end_of_text|>",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend"
|
| 15 |
+
}
|
checkpoint-534/trainer_state.json
ADDED
|
@@ -0,0 +1,1708 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 534,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0,
|
| 14 |
+
"forget_loss": 0.00946044921875,
|
| 15 |
+
"retain_loss": 0.0,
|
| 16 |
+
"step": 0
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0,
|
| 20 |
+
"forget_loss": 0.00994873046875,
|
| 21 |
+
"retain_loss": 0.0,
|
| 22 |
+
"step": 0
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 0,
|
| 26 |
+
"forget_loss": 0.00958251953125,
|
| 27 |
+
"retain_loss": 0.0,
|
| 28 |
+
"step": 0
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0,
|
| 32 |
+
"forget_loss": 0.009521484375,
|
| 33 |
+
"retain_loss": 0.0,
|
| 34 |
+
"step": 0
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.003755868544600939,
|
| 38 |
+
"grad_norm": 0.11474609375,
|
| 39 |
+
"learning_rate": 0.0,
|
| 40 |
+
"loss": 0.0096282958984375,
|
| 41 |
+
"step": 1
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.03755868544600939,
|
| 45 |
+
"grad_norm": 0.09814453125,
|
| 46 |
+
"learning_rate": 4.5e-06,
|
| 47 |
+
"loss": 0.009236653645833334,
|
| 48 |
+
"step": 10
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.03755868544600939,
|
| 52 |
+
"forget_loss": 0.00799560546875,
|
| 53 |
+
"retain_loss": 1.919269561767578e-05,
|
| 54 |
+
"step": 10
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 0.03755868544600939,
|
| 58 |
+
"forget_loss": 0.008056640625,
|
| 59 |
+
"retain_loss": 1.6570091247558594e-05,
|
| 60 |
+
"step": 10
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.03755868544600939,
|
| 64 |
+
"forget_loss": 0.0081787109375,
|
| 65 |
+
"retain_loss": 2.2649765014648438e-05,
|
| 66 |
+
"step": 10
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.03755868544600939,
|
| 70 |
+
"forget_loss": 0.00823974609375,
|
| 71 |
+
"retain_loss": 2.1457672119140625e-05,
|
| 72 |
+
"step": 10
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.07511737089201878,
|
| 76 |
+
"grad_norm": 0.047607421875,
|
| 77 |
+
"learning_rate": 9.5e-06,
|
| 78 |
+
"loss": 0.00635223388671875,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.07511737089201878,
|
| 83 |
+
"forget_loss": 0.003662109375,
|
| 84 |
+
"retain_loss": 0.00022602081298828125,
|
| 85 |
+
"step": 20
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.07511737089201878,
|
| 89 |
+
"forget_loss": 0.004425048828125,
|
| 90 |
+
"retain_loss": 0.0002880096435546875,
|
| 91 |
+
"step": 20
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 0.07511737089201878,
|
| 95 |
+
"forget_loss": 0.0037078857421875,
|
| 96 |
+
"retain_loss": 0.00020885467529296875,
|
| 97 |
+
"step": 20
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.07511737089201878,
|
| 101 |
+
"forget_loss": 0.00421142578125,
|
| 102 |
+
"retain_loss": 0.00021457672119140625,
|
| 103 |
+
"step": 20
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.11267605633802817,
|
| 107 |
+
"grad_norm": 0.123046875,
|
| 108 |
+
"learning_rate": 9.82490272373541e-06,
|
| 109 |
+
"loss": 0.00320587158203125,
|
| 110 |
+
"step": 30
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.11267605633802817,
|
| 114 |
+
"forget_loss": 0.0018310546875,
|
| 115 |
+
"retain_loss": 0.00054168701171875,
|
| 116 |
+
"step": 30
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.11267605633802817,
|
| 120 |
+
"forget_loss": 0.00177001953125,
|
| 121 |
+
"retain_loss": 0.0004177093505859375,
|
| 122 |
+
"step": 30
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.11267605633802817,
|
| 126 |
+
"forget_loss": 0.00177001953125,
|
| 127 |
+
"retain_loss": 0.0002956390380859375,
|
| 128 |
+
"step": 30
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.11267605633802817,
|
| 132 |
+
"forget_loss": 0.0016937255859375,
|
| 133 |
+
"retain_loss": 0.000331878662109375,
|
| 134 |
+
"step": 30
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.15023474178403756,
|
| 138 |
+
"grad_norm": 0.05517578125,
|
| 139 |
+
"learning_rate": 9.63035019455253e-06,
|
| 140 |
+
"loss": 0.0016145706176757812,
|
| 141 |
+
"step": 40
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 0.15023474178403756,
|
| 145 |
+
"forget_loss": 0.0007781982421875,
|
| 146 |
+
"retain_loss": 0.00021648406982421875,
|
| 147 |
+
"step": 40
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.15023474178403756,
|
| 151 |
+
"forget_loss": 0.000762939453125,
|
| 152 |
+
"retain_loss": 0.0004100799560546875,
|
| 153 |
+
"step": 40
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.15023474178403756,
|
| 157 |
+
"forget_loss": 0.000766754150390625,
|
| 158 |
+
"retain_loss": 0.000331878662109375,
|
| 159 |
+
"step": 40
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 0.15023474178403756,
|
| 163 |
+
"forget_loss": 0.0007781982421875,
|
| 164 |
+
"retain_loss": 0.00038909912109375,
|
| 165 |
+
"step": 40
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 0.18779342723004694,
|
| 169 |
+
"grad_norm": 0.1416015625,
|
| 170 |
+
"learning_rate": 9.435797665369651e-06,
|
| 171 |
+
"loss": 0.000954437255859375,
|
| 172 |
+
"step": 50
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.18779342723004694,
|
| 176 |
+
"forget_loss": 0.0003452301025390625,
|
| 177 |
+
"retain_loss": 0.00020599365234375,
|
| 178 |
+
"step": 50
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.18779342723004694,
|
| 182 |
+
"forget_loss": 0.0003814697265625,
|
| 183 |
+
"retain_loss": 0.00020694732666015625,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.18779342723004694,
|
| 188 |
+
"forget_loss": 0.000415802001953125,
|
| 189 |
+
"retain_loss": 0.0003604888916015625,
|
| 190 |
+
"step": 50
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.18779342723004694,
|
| 194 |
+
"forget_loss": 0.0003719329833984375,
|
| 195 |
+
"retain_loss": 0.0001964569091796875,
|
| 196 |
+
"step": 50
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.22535211267605634,
|
| 200 |
+
"grad_norm": 0.06396484375,
|
| 201 |
+
"learning_rate": 9.241245136186772e-06,
|
| 202 |
+
"loss": 0.0008329868316650391,
|
| 203 |
+
"step": 60
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.22535211267605634,
|
| 207 |
+
"forget_loss": 0.0002574920654296875,
|
| 208 |
+
"retain_loss": 0.0002727508544921875,
|
| 209 |
+
"step": 60
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.22535211267605634,
|
| 213 |
+
"forget_loss": 0.00019931793212890625,
|
| 214 |
+
"retain_loss": 0.00010919570922851562,
|
| 215 |
+
"step": 60
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.22535211267605634,
|
| 219 |
+
"forget_loss": 0.0001983642578125,
|
| 220 |
+
"retain_loss": 0.000202178955078125,
|
| 221 |
+
"step": 60
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.22535211267605634,
|
| 225 |
+
"forget_loss": 0.00023365020751953125,
|
| 226 |
+
"retain_loss": 0.0002002716064453125,
|
| 227 |
+
"step": 60
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.26291079812206575,
|
| 231 |
+
"grad_norm": 0.28515625,
|
| 232 |
+
"learning_rate": 9.046692607003892e-06,
|
| 233 |
+
"loss": 0.0006081581115722657,
|
| 234 |
+
"step": 70
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.26291079812206575,
|
| 238 |
+
"forget_loss": 0.00018405914306640625,
|
| 239 |
+
"retain_loss": 0.0002689361572265625,
|
| 240 |
+
"step": 70
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.26291079812206575,
|
| 244 |
+
"forget_loss": 0.0001811981201171875,
|
| 245 |
+
"retain_loss": 8.535385131835938e-05,
|
| 246 |
+
"step": 70
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.26291079812206575,
|
| 250 |
+
"forget_loss": 0.00022792816162109375,
|
| 251 |
+
"retain_loss": 0.00010728836059570312,
|
| 252 |
+
"step": 70
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 0.26291079812206575,
|
| 256 |
+
"forget_loss": 0.0003643035888671875,
|
| 257 |
+
"retain_loss": 0.00011777877807617188,
|
| 258 |
+
"step": 70
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.3004694835680751,
|
| 262 |
+
"grad_norm": 0.07177734375,
|
| 263 |
+
"learning_rate": 8.852140077821012e-06,
|
| 264 |
+
"loss": 0.0004550933837890625,
|
| 265 |
+
"step": 80
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.3004694835680751,
|
| 269 |
+
"forget_loss": 0.0002155303955078125,
|
| 270 |
+
"retain_loss": 0.000110626220703125,
|
| 271 |
+
"step": 80
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.3004694835680751,
|
| 275 |
+
"forget_loss": 0.00014019012451171875,
|
| 276 |
+
"retain_loss": 6.723403930664062e-05,
|
| 277 |
+
"step": 80
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 0.3004694835680751,
|
| 281 |
+
"forget_loss": 0.0001773834228515625,
|
| 282 |
+
"retain_loss": 0.000331878662109375,
|
| 283 |
+
"step": 80
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.3004694835680751,
|
| 287 |
+
"forget_loss": 0.00016498565673828125,
|
| 288 |
+
"retain_loss": 9.489059448242188e-05,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.3380281690140845,
|
| 293 |
+
"grad_norm": 0.0810546875,
|
| 294 |
+
"learning_rate": 8.657587548638134e-06,
|
| 295 |
+
"loss": 0.0003980875015258789,
|
| 296 |
+
"step": 90
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.3380281690140845,
|
| 300 |
+
"forget_loss": 0.00015735626220703125,
|
| 301 |
+
"retain_loss": 0.0001544952392578125,
|
| 302 |
+
"step": 90
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.3380281690140845,
|
| 306 |
+
"forget_loss": 0.000133514404296875,
|
| 307 |
+
"retain_loss": 0.00080108642578125,
|
| 308 |
+
"step": 90
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.3380281690140845,
|
| 312 |
+
"forget_loss": 0.00014781951904296875,
|
| 313 |
+
"retain_loss": 0.000431060791015625,
|
| 314 |
+
"step": 90
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.3380281690140845,
|
| 318 |
+
"forget_loss": 0.00014400482177734375,
|
| 319 |
+
"retain_loss": 0.00031280517578125,
|
| 320 |
+
"step": 90
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.3755868544600939,
|
| 324 |
+
"grad_norm": 0.025146484375,
|
| 325 |
+
"learning_rate": 8.463035019455253e-06,
|
| 326 |
+
"loss": 0.00032124519348144533,
|
| 327 |
+
"step": 100
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.3755868544600939,
|
| 331 |
+
"forget_loss": 0.00015163421630859375,
|
| 332 |
+
"retain_loss": 0.00015926361083984375,
|
| 333 |
+
"step": 100
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 0.3755868544600939,
|
| 337 |
+
"forget_loss": 0.0001506805419921875,
|
| 338 |
+
"retain_loss": 0.00011587142944335938,
|
| 339 |
+
"step": 100
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.3755868544600939,
|
| 343 |
+
"forget_loss": 0.00013065338134765625,
|
| 344 |
+
"retain_loss": 0.000125885009765625,
|
| 345 |
+
"step": 100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3755868544600939,
|
| 349 |
+
"forget_loss": 0.0001239776611328125,
|
| 350 |
+
"retain_loss": 0.0001277923583984375,
|
| 351 |
+
"step": 100
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.4131455399061033,
|
| 355 |
+
"grad_norm": 0.06884765625,
|
| 356 |
+
"learning_rate": 8.268482490272373e-06,
|
| 357 |
+
"loss": 0.00034072399139404295,
|
| 358 |
+
"step": 110
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.4131455399061033,
|
| 362 |
+
"forget_loss": 0.0001544952392578125,
|
| 363 |
+
"retain_loss": 7.2479248046875e-05,
|
| 364 |
+
"step": 110
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.4131455399061033,
|
| 368 |
+
"forget_loss": 0.0001373291015625,
|
| 369 |
+
"retain_loss": 0.000270843505859375,
|
| 370 |
+
"step": 110
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.4131455399061033,
|
| 374 |
+
"forget_loss": 0.0001125335693359375,
|
| 375 |
+
"retain_loss": 9.965896606445312e-05,
|
| 376 |
+
"step": 110
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.4131455399061033,
|
| 380 |
+
"forget_loss": 0.00011205673217773438,
|
| 381 |
+
"retain_loss": 6.866455078125e-05,
|
| 382 |
+
"step": 110
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.4507042253521127,
|
| 386 |
+
"grad_norm": 0.08544921875,
|
| 387 |
+
"learning_rate": 8.073929961089495e-06,
|
| 388 |
+
"loss": 0.0002372264862060547,
|
| 389 |
+
"step": 120
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.4507042253521127,
|
| 393 |
+
"forget_loss": 0.000179290771484375,
|
| 394 |
+
"retain_loss": 7.05718994140625e-05,
|
| 395 |
+
"step": 120
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.4507042253521127,
|
| 399 |
+
"forget_loss": 0.0001277923583984375,
|
| 400 |
+
"retain_loss": 9.250640869140625e-05,
|
| 401 |
+
"step": 120
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.4507042253521127,
|
| 405 |
+
"forget_loss": 0.00013446807861328125,
|
| 406 |
+
"retain_loss": 8.630752563476562e-05,
|
| 407 |
+
"step": 120
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.4507042253521127,
|
| 411 |
+
"forget_loss": 0.0001735687255859375,
|
| 412 |
+
"retain_loss": 8.106231689453125e-05,
|
| 413 |
+
"step": 120
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 0.48826291079812206,
|
| 417 |
+
"grad_norm": 0.0361328125,
|
| 418 |
+
"learning_rate": 7.879377431906614e-06,
|
| 419 |
+
"loss": 0.0002105236053466797,
|
| 420 |
+
"step": 130
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 0.48826291079812206,
|
| 424 |
+
"forget_loss": 0.00010776519775390625,
|
| 425 |
+
"retain_loss": 5.221366882324219e-05,
|
| 426 |
+
"step": 130
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.48826291079812206,
|
| 430 |
+
"forget_loss": 0.00010728836059570312,
|
| 431 |
+
"retain_loss": 5.1021575927734375e-05,
|
| 432 |
+
"step": 130
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.48826291079812206,
|
| 436 |
+
"forget_loss": 0.00010204315185546875,
|
| 437 |
+
"retain_loss": 0.00011110305786132812,
|
| 438 |
+
"step": 130
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.48826291079812206,
|
| 442 |
+
"forget_loss": 0.000156402587890625,
|
| 443 |
+
"retain_loss": 7.05718994140625e-05,
|
| 444 |
+
"step": 130
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.5258215962441315,
|
| 448 |
+
"grad_norm": 0.035400390625,
|
| 449 |
+
"learning_rate": 7.684824902723736e-06,
|
| 450 |
+
"loss": 0.000272369384765625,
|
| 451 |
+
"step": 140
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.5258215962441315,
|
| 455 |
+
"forget_loss": 0.0001087188720703125,
|
| 456 |
+
"retain_loss": 7.009506225585938e-05,
|
| 457 |
+
"step": 140
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.5258215962441315,
|
| 461 |
+
"forget_loss": 0.0001239776611328125,
|
| 462 |
+
"retain_loss": 7.963180541992188e-05,
|
| 463 |
+
"step": 140
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.5258215962441315,
|
| 467 |
+
"forget_loss": 0.00017070770263671875,
|
| 468 |
+
"retain_loss": 6.29425048828125e-05,
|
| 469 |
+
"step": 140
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.5258215962441315,
|
| 473 |
+
"forget_loss": 0.00013446807861328125,
|
| 474 |
+
"retain_loss": 5.5789947509765625e-05,
|
| 475 |
+
"step": 140
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.5633802816901409,
|
| 479 |
+
"grad_norm": 0.0177001953125,
|
| 480 |
+
"learning_rate": 7.4902723735408565e-06,
|
| 481 |
+
"loss": 0.00019230842590332032,
|
| 482 |
+
"step": 150
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.5633802816901409,
|
| 486 |
+
"forget_loss": 0.00010585784912109375,
|
| 487 |
+
"retain_loss": 5.793571472167969e-05,
|
| 488 |
+
"step": 150
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.5633802816901409,
|
| 492 |
+
"forget_loss": 0.00011157989501953125,
|
| 493 |
+
"retain_loss": 0.00015163421630859375,
|
| 494 |
+
"step": 150
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.5633802816901409,
|
| 498 |
+
"forget_loss": 8.821487426757812e-05,
|
| 499 |
+
"retain_loss": 6.866455078125e-05,
|
| 500 |
+
"step": 150
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.5633802816901409,
|
| 504 |
+
"forget_loss": 9.059906005859375e-05,
|
| 505 |
+
"retain_loss": 0.00010251998901367188,
|
| 506 |
+
"step": 150
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.6009389671361502,
|
| 510 |
+
"grad_norm": 0.06591796875,
|
| 511 |
+
"learning_rate": 7.295719844357977e-06,
|
| 512 |
+
"loss": 0.00020923614501953126,
|
| 513 |
+
"step": 160
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.6009389671361502,
|
| 517 |
+
"forget_loss": 0.00010156631469726562,
|
| 518 |
+
"retain_loss": 8.535385131835938e-05,
|
| 519 |
+
"step": 160
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.6009389671361502,
|
| 523 |
+
"forget_loss": 9.441375732421875e-05,
|
| 524 |
+
"retain_loss": 7.05718994140625e-05,
|
| 525 |
+
"step": 160
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.6009389671361502,
|
| 529 |
+
"forget_loss": 9.489059448242188e-05,
|
| 530 |
+
"retain_loss": 6.151199340820312e-05,
|
| 531 |
+
"step": 160
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.6009389671361502,
|
| 535 |
+
"forget_loss": 9.679794311523438e-05,
|
| 536 |
+
"retain_loss": 5.984306335449219e-05,
|
| 537 |
+
"step": 160
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.6384976525821596,
|
| 541 |
+
"grad_norm": 0.0152587890625,
|
| 542 |
+
"learning_rate": 7.101167315175098e-06,
|
| 543 |
+
"loss": 0.0001931428909301758,
|
| 544 |
+
"step": 170
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.6384976525821596,
|
| 548 |
+
"forget_loss": 9.489059448242188e-05,
|
| 549 |
+
"retain_loss": 6.079673767089844e-05,
|
| 550 |
+
"step": 170
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.6384976525821596,
|
| 554 |
+
"forget_loss": 0.000110626220703125,
|
| 555 |
+
"retain_loss": 7.104873657226562e-05,
|
| 556 |
+
"step": 170
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.6384976525821596,
|
| 560 |
+
"forget_loss": 0.00010776519775390625,
|
| 561 |
+
"retain_loss": 0.00011539459228515625,
|
| 562 |
+
"step": 170
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.6384976525821596,
|
| 566 |
+
"forget_loss": 9.441375732421875e-05,
|
| 567 |
+
"retain_loss": 0.00011682510375976562,
|
| 568 |
+
"step": 170
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.676056338028169,
|
| 572 |
+
"grad_norm": 0.0306396484375,
|
| 573 |
+
"learning_rate": 6.9066147859922184e-06,
|
| 574 |
+
"loss": 0.00021104812622070314,
|
| 575 |
+
"step": 180
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 0.676056338028169,
|
| 579 |
+
"forget_loss": 0.000125885009765625,
|
| 580 |
+
"retain_loss": 0.0001239776611328125,
|
| 581 |
+
"step": 180
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 0.676056338028169,
|
| 585 |
+
"forget_loss": 0.00010395050048828125,
|
| 586 |
+
"retain_loss": 8.96453857421875e-05,
|
| 587 |
+
"step": 180
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 0.676056338028169,
|
| 591 |
+
"forget_loss": 8.869171142578125e-05,
|
| 592 |
+
"retain_loss": 8.20159912109375e-05,
|
| 593 |
+
"step": 180
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.676056338028169,
|
| 597 |
+
"forget_loss": 9.107589721679688e-05,
|
| 598 |
+
"retain_loss": 5.316734313964844e-05,
|
| 599 |
+
"step": 180
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.7136150234741784,
|
| 603 |
+
"grad_norm": 0.0230712890625,
|
| 604 |
+
"learning_rate": 6.712062256809338e-06,
|
| 605 |
+
"loss": 0.0001739978790283203,
|
| 606 |
+
"step": 190
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.7136150234741784,
|
| 610 |
+
"forget_loss": 9.059906005859375e-05,
|
| 611 |
+
"retain_loss": 5.793571472167969e-05,
|
| 612 |
+
"step": 190
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.7136150234741784,
|
| 616 |
+
"forget_loss": 9.5367431640625e-05,
|
| 617 |
+
"retain_loss": 5.0067901611328125e-05,
|
| 618 |
+
"step": 190
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.7136150234741784,
|
| 622 |
+
"forget_loss": 8.869171142578125e-05,
|
| 623 |
+
"retain_loss": 5.7220458984375e-05,
|
| 624 |
+
"step": 190
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.7136150234741784,
|
| 628 |
+
"forget_loss": 0.00010919570922851562,
|
| 629 |
+
"retain_loss": 4.9591064453125e-05,
|
| 630 |
+
"step": 190
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.7511737089201878,
|
| 634 |
+
"grad_norm": 0.01068115234375,
|
| 635 |
+
"learning_rate": 6.51750972762646e-06,
|
| 636 |
+
"loss": 0.00019364356994628907,
|
| 637 |
+
"step": 200
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 0.7511737089201878,
|
| 641 |
+
"forget_loss": 0.00010204315185546875,
|
| 642 |
+
"retain_loss": 5.14984130859375e-05,
|
| 643 |
+
"step": 200
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.7511737089201878,
|
| 647 |
+
"forget_loss": 8.249282836914062e-05,
|
| 648 |
+
"retain_loss": 4.887580871582031e-05,
|
| 649 |
+
"step": 200
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.7511737089201878,
|
| 653 |
+
"forget_loss": 7.772445678710938e-05,
|
| 654 |
+
"retain_loss": 4.076957702636719e-05,
|
| 655 |
+
"step": 200
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.7511737089201878,
|
| 659 |
+
"forget_loss": 0.00012683868408203125,
|
| 660 |
+
"retain_loss": 6.341934204101562e-05,
|
| 661 |
+
"step": 200
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.7887323943661971,
|
| 665 |
+
"grad_norm": 0.016357421875,
|
| 666 |
+
"learning_rate": 6.32295719844358e-06,
|
| 667 |
+
"loss": 0.00015709400177001953,
|
| 668 |
+
"step": 210
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.7887323943661971,
|
| 672 |
+
"forget_loss": 0.00010585784912109375,
|
| 673 |
+
"retain_loss": 7.009506225585938e-05,
|
| 674 |
+
"step": 210
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.7887323943661971,
|
| 678 |
+
"forget_loss": 8.58306884765625e-05,
|
| 679 |
+
"retain_loss": 4.2438507080078125e-05,
|
| 680 |
+
"step": 210
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.7887323943661971,
|
| 684 |
+
"forget_loss": 8.440017700195312e-05,
|
| 685 |
+
"retain_loss": 4.744529724121094e-05,
|
| 686 |
+
"step": 210
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.7887323943661971,
|
| 690 |
+
"forget_loss": 0.0001010894775390625,
|
| 691 |
+
"retain_loss": 4.4345855712890625e-05,
|
| 692 |
+
"step": 210
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.8262910798122066,
|
| 696 |
+
"grad_norm": 0.01806640625,
|
| 697 |
+
"learning_rate": 6.128404669260701e-06,
|
| 698 |
+
"loss": 0.0001538991928100586,
|
| 699 |
+
"step": 220
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.8262910798122066,
|
| 703 |
+
"forget_loss": 9.72747802734375e-05,
|
| 704 |
+
"retain_loss": 4.8160552978515625e-05,
|
| 705 |
+
"step": 220
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.8262910798122066,
|
| 709 |
+
"forget_loss": 8.0108642578125e-05,
|
| 710 |
+
"retain_loss": 7.200241088867188e-05,
|
| 711 |
+
"step": 220
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.8262910798122066,
|
| 715 |
+
"forget_loss": 7.82012939453125e-05,
|
| 716 |
+
"retain_loss": 4.3392181396484375e-05,
|
| 717 |
+
"step": 220
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.8262910798122066,
|
| 721 |
+
"forget_loss": 0.0001049041748046875,
|
| 722 |
+
"retain_loss": 7.009506225585938e-05,
|
| 723 |
+
"step": 220
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.863849765258216,
|
| 727 |
+
"grad_norm": 0.0184326171875,
|
| 728 |
+
"learning_rate": 5.933852140077822e-06,
|
| 729 |
+
"loss": 0.00015753507614135742,
|
| 730 |
+
"step": 230
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.863849765258216,
|
| 734 |
+
"forget_loss": 8.678436279296875e-05,
|
| 735 |
+
"retain_loss": 4.38690185546875e-05,
|
| 736 |
+
"step": 230
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.863849765258216,
|
| 740 |
+
"forget_loss": 8.344650268554688e-05,
|
| 741 |
+
"retain_loss": 0.000942230224609375,
|
| 742 |
+
"step": 230
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.863849765258216,
|
| 746 |
+
"forget_loss": 9.1552734375e-05,
|
| 747 |
+
"retain_loss": 4.410743713378906e-05,
|
| 748 |
+
"step": 230
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 0.863849765258216,
|
| 752 |
+
"forget_loss": 8.296966552734375e-05,
|
| 753 |
+
"retain_loss": 7.486343383789062e-05,
|
| 754 |
+
"step": 230
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.9014084507042254,
|
| 758 |
+
"grad_norm": 0.0238037109375,
|
| 759 |
+
"learning_rate": 5.7392996108949415e-06,
|
| 760 |
+
"loss": 0.00017303228378295898,
|
| 761 |
+
"step": 240
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.9014084507042254,
|
| 765 |
+
"forget_loss": 8.296966552734375e-05,
|
| 766 |
+
"retain_loss": 4.291534423828125e-05,
|
| 767 |
+
"step": 240
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.9014084507042254,
|
| 771 |
+
"forget_loss": 8.96453857421875e-05,
|
| 772 |
+
"retain_loss": 3.814697265625e-05,
|
| 773 |
+
"step": 240
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 0.9014084507042254,
|
| 777 |
+
"forget_loss": 8.726119995117188e-05,
|
| 778 |
+
"retain_loss": 4.839897155761719e-05,
|
| 779 |
+
"step": 240
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.9014084507042254,
|
| 783 |
+
"forget_loss": 7.200241088867188e-05,
|
| 784 |
+
"retain_loss": 4.839897155761719e-05,
|
| 785 |
+
"step": 240
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.9389671361502347,
|
| 789 |
+
"grad_norm": 0.05908203125,
|
| 790 |
+
"learning_rate": 5.544747081712063e-06,
|
| 791 |
+
"loss": 0.00014340877532958984,
|
| 792 |
+
"step": 250
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.9389671361502347,
|
| 796 |
+
"forget_loss": 0.00010442733764648438,
|
| 797 |
+
"retain_loss": 5.626678466796875e-05,
|
| 798 |
+
"step": 250
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.9389671361502347,
|
| 802 |
+
"forget_loss": 8.726119995117188e-05,
|
| 803 |
+
"retain_loss": 9.059906005859375e-05,
|
| 804 |
+
"step": 250
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.9389671361502347,
|
| 808 |
+
"forget_loss": 8.344650268554688e-05,
|
| 809 |
+
"retain_loss": 3.647804260253906e-05,
|
| 810 |
+
"step": 250
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.9389671361502347,
|
| 814 |
+
"forget_loss": 8.821487426757812e-05,
|
| 815 |
+
"retain_loss": 4.220008850097656e-05,
|
| 816 |
+
"step": 250
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.9765258215962441,
|
| 820 |
+
"grad_norm": 0.0181884765625,
|
| 821 |
+
"learning_rate": 5.350194552529184e-06,
|
| 822 |
+
"loss": 0.00015048980712890626,
|
| 823 |
+
"step": 260
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.9765258215962441,
|
| 827 |
+
"forget_loss": 8.153915405273438e-05,
|
| 828 |
+
"retain_loss": 6.580352783203125e-05,
|
| 829 |
+
"step": 260
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 0.9765258215962441,
|
| 833 |
+
"forget_loss": 8.7738037109375e-05,
|
| 834 |
+
"retain_loss": 3.6716461181640625e-05,
|
| 835 |
+
"step": 260
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.9765258215962441,
|
| 839 |
+
"forget_loss": 9.918212890625e-05,
|
| 840 |
+
"retain_loss": 4.363059997558594e-05,
|
| 841 |
+
"step": 260
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.9765258215962441,
|
| 845 |
+
"forget_loss": 9.870529174804688e-05,
|
| 846 |
+
"retain_loss": 0.0001506805419921875,
|
| 847 |
+
"step": 260
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 1.0112676056338028,
|
| 851 |
+
"grad_norm": 0.00848388671875,
|
| 852 |
+
"learning_rate": 5.155642023346303e-06,
|
| 853 |
+
"loss": 0.00013761520385742189,
|
| 854 |
+
"step": 270
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.0112676056338028,
|
| 858 |
+
"forget_loss": 6.866455078125e-05,
|
| 859 |
+
"retain_loss": 5.9604644775390625e-05,
|
| 860 |
+
"step": 270
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 1.0112676056338028,
|
| 864 |
+
"forget_loss": 8.249282836914062e-05,
|
| 865 |
+
"retain_loss": 3.5762786865234375e-05,
|
| 866 |
+
"step": 270
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.0112676056338028,
|
| 870 |
+
"forget_loss": 8.392333984375e-05,
|
| 871 |
+
"retain_loss": 6.246566772460938e-05,
|
| 872 |
+
"step": 270
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.0112676056338028,
|
| 876 |
+
"forget_loss": 8.916854858398438e-05,
|
| 877 |
+
"retain_loss": 3.695487976074219e-05,
|
| 878 |
+
"step": 270
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.0488262910798123,
|
| 882 |
+
"grad_norm": 0.0137939453125,
|
| 883 |
+
"learning_rate": 4.961089494163425e-06,
|
| 884 |
+
"loss": 0.00013872385025024415,
|
| 885 |
+
"step": 280
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.0488262910798123,
|
| 889 |
+
"forget_loss": 7.772445678710938e-05,
|
| 890 |
+
"retain_loss": 4.982948303222656e-05,
|
| 891 |
+
"step": 280
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.0488262910798123,
|
| 895 |
+
"forget_loss": 6.771087646484375e-05,
|
| 896 |
+
"retain_loss": 4.5299530029296875e-05,
|
| 897 |
+
"step": 280
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 1.0488262910798123,
|
| 901 |
+
"forget_loss": 6.961822509765625e-05,
|
| 902 |
+
"retain_loss": 5.14984130859375e-05,
|
| 903 |
+
"step": 280
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 1.0488262910798123,
|
| 907 |
+
"forget_loss": 7.05718994140625e-05,
|
| 908 |
+
"retain_loss": 5.245208740234375e-05,
|
| 909 |
+
"step": 280
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 1.0863849765258216,
|
| 913 |
+
"grad_norm": 0.01275634765625,
|
| 914 |
+
"learning_rate": 4.7665369649805455e-06,
|
| 915 |
+
"loss": 0.00014275312423706055,
|
| 916 |
+
"step": 290
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 1.0863849765258216,
|
| 920 |
+
"forget_loss": 8.20159912109375e-05,
|
| 921 |
+
"retain_loss": 4.38690185546875e-05,
|
| 922 |
+
"step": 290
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 1.0863849765258216,
|
| 926 |
+
"forget_loss": 7.62939453125e-05,
|
| 927 |
+
"retain_loss": 5.745887756347656e-05,
|
| 928 |
+
"step": 290
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.0863849765258216,
|
| 932 |
+
"forget_loss": 9.059906005859375e-05,
|
| 933 |
+
"retain_loss": 5.269050598144531e-05,
|
| 934 |
+
"step": 290
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.0863849765258216,
|
| 938 |
+
"forget_loss": 8.106231689453125e-05,
|
| 939 |
+
"retain_loss": 5.507469177246094e-05,
|
| 940 |
+
"step": 290
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.123943661971831,
|
| 944 |
+
"grad_norm": 0.0140380859375,
|
| 945 |
+
"learning_rate": 4.571984435797665e-06,
|
| 946 |
+
"loss": 0.00013855695724487304,
|
| 947 |
+
"step": 300
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.123943661971831,
|
| 951 |
+
"forget_loss": 8.344650268554688e-05,
|
| 952 |
+
"retain_loss": 3.695487976074219e-05,
|
| 953 |
+
"step": 300
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 1.123943661971831,
|
| 957 |
+
"forget_loss": 8.249282836914062e-05,
|
| 958 |
+
"retain_loss": 4.100799560546875e-05,
|
| 959 |
+
"step": 300
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 1.123943661971831,
|
| 963 |
+
"forget_loss": 0.00010395050048828125,
|
| 964 |
+
"retain_loss": 3.504753112792969e-05,
|
| 965 |
+
"step": 300
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 1.123943661971831,
|
| 969 |
+
"forget_loss": 6.4849853515625e-05,
|
| 970 |
+
"retain_loss": 5.5789947509765625e-05,
|
| 971 |
+
"step": 300
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.1615023474178403,
|
| 975 |
+
"grad_norm": 0.019775390625,
|
| 976 |
+
"learning_rate": 4.377431906614786e-06,
|
| 977 |
+
"loss": 0.0001352071762084961,
|
| 978 |
+
"step": 310
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 1.1615023474178403,
|
| 982 |
+
"forget_loss": 7.62939453125e-05,
|
| 983 |
+
"retain_loss": 7.772445678710938e-05,
|
| 984 |
+
"step": 310
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.1615023474178403,
|
| 988 |
+
"forget_loss": 8.58306884765625e-05,
|
| 989 |
+
"retain_loss": 5.459785461425781e-05,
|
| 990 |
+
"step": 310
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.1615023474178403,
|
| 994 |
+
"forget_loss": 7.05718994140625e-05,
|
| 995 |
+
"retain_loss": 7.05718994140625e-05,
|
| 996 |
+
"step": 310
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.1615023474178403,
|
| 1000 |
+
"forget_loss": 7.82012939453125e-05,
|
| 1001 |
+
"retain_loss": 3.337860107421875e-05,
|
| 1002 |
+
"step": 310
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 1.1990610328638498,
|
| 1006 |
+
"grad_norm": 0.020751953125,
|
| 1007 |
+
"learning_rate": 4.1828793774319074e-06,
|
| 1008 |
+
"loss": 0.000127410888671875,
|
| 1009 |
+
"step": 320
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 1.1990610328638498,
|
| 1013 |
+
"forget_loss": 7.867813110351562e-05,
|
| 1014 |
+
"retain_loss": 3.695487976074219e-05,
|
| 1015 |
+
"step": 320
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 1.1990610328638498,
|
| 1019 |
+
"forget_loss": 8.249282836914062e-05,
|
| 1020 |
+
"retain_loss": 5.412101745605469e-05,
|
| 1021 |
+
"step": 320
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 1.1990610328638498,
|
| 1025 |
+
"forget_loss": 9.775161743164062e-05,
|
| 1026 |
+
"retain_loss": 5.340576171875e-05,
|
| 1027 |
+
"step": 320
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 1.1990610328638498,
|
| 1031 |
+
"forget_loss": 7.152557373046875e-05,
|
| 1032 |
+
"retain_loss": 8.106231689453125e-05,
|
| 1033 |
+
"step": 320
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 1.236619718309859,
|
| 1037 |
+
"grad_norm": 0.0091552734375,
|
| 1038 |
+
"learning_rate": 3.988326848249028e-06,
|
| 1039 |
+
"loss": 0.00014822483062744142,
|
| 1040 |
+
"step": 330
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 1.236619718309859,
|
| 1044 |
+
"forget_loss": 9.72747802734375e-05,
|
| 1045 |
+
"retain_loss": 4.506111145019531e-05,
|
| 1046 |
+
"step": 330
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.236619718309859,
|
| 1050 |
+
"forget_loss": 0.00010919570922851562,
|
| 1051 |
+
"retain_loss": 3.2901763916015625e-05,
|
| 1052 |
+
"step": 330
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.236619718309859,
|
| 1056 |
+
"forget_loss": 6.723403930664062e-05,
|
| 1057 |
+
"retain_loss": 4.029273986816406e-05,
|
| 1058 |
+
"step": 330
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 1.236619718309859,
|
| 1062 |
+
"forget_loss": 6.914138793945312e-05,
|
| 1063 |
+
"retain_loss": 7.724761962890625e-05,
|
| 1064 |
+
"step": 330
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 1.2741784037558685,
|
| 1068 |
+
"grad_norm": 0.011962890625,
|
| 1069 |
+
"learning_rate": 3.793774319066148e-06,
|
| 1070 |
+
"loss": 0.0001390695571899414,
|
| 1071 |
+
"step": 340
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 1.2741784037558685,
|
| 1075 |
+
"forget_loss": 8.535385131835938e-05,
|
| 1076 |
+
"retain_loss": 4.506111145019531e-05,
|
| 1077 |
+
"step": 340
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 1.2741784037558685,
|
| 1081 |
+
"forget_loss": 6.866455078125e-05,
|
| 1082 |
+
"retain_loss": 3.2901763916015625e-05,
|
| 1083 |
+
"step": 340
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 1.2741784037558685,
|
| 1087 |
+
"forget_loss": 6.723403930664062e-05,
|
| 1088 |
+
"retain_loss": 5.0067901611328125e-05,
|
| 1089 |
+
"step": 340
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.2741784037558685,
|
| 1093 |
+
"forget_loss": 8.344650268554688e-05,
|
| 1094 |
+
"retain_loss": 3.5762786865234375e-05,
|
| 1095 |
+
"step": 340
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.3117370892018778,
|
| 1099 |
+
"grad_norm": 0.01239013671875,
|
| 1100 |
+
"learning_rate": 3.599221789883269e-06,
|
| 1101 |
+
"loss": 0.00013332366943359376,
|
| 1102 |
+
"step": 350
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 1.3117370892018778,
|
| 1106 |
+
"forget_loss": 7.772445678710938e-05,
|
| 1107 |
+
"retain_loss": 3.0040740966796875e-05,
|
| 1108 |
+
"step": 350
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.3117370892018778,
|
| 1112 |
+
"forget_loss": 7.963180541992188e-05,
|
| 1113 |
+
"retain_loss": 4.363059997558594e-05,
|
| 1114 |
+
"step": 350
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 1.3117370892018778,
|
| 1118 |
+
"forget_loss": 8.630752563476562e-05,
|
| 1119 |
+
"retain_loss": 4.6253204345703125e-05,
|
| 1120 |
+
"step": 350
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 1.3117370892018778,
|
| 1124 |
+
"forget_loss": 8.487701416015625e-05,
|
| 1125 |
+
"retain_loss": 5.030632019042969e-05,
|
| 1126 |
+
"step": 350
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 1.3492957746478873,
|
| 1130 |
+
"grad_norm": 0.01177978515625,
|
| 1131 |
+
"learning_rate": 3.4046692607003896e-06,
|
| 1132 |
+
"loss": 0.00012853145599365235,
|
| 1133 |
+
"step": 360
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 1.3492957746478873,
|
| 1137 |
+
"forget_loss": 0.00010395050048828125,
|
| 1138 |
+
"retain_loss": 3.3855438232421875e-05,
|
| 1139 |
+
"step": 360
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 1.3492957746478873,
|
| 1143 |
+
"forget_loss": 9.250640869140625e-05,
|
| 1144 |
+
"retain_loss": 4.553794860839844e-05,
|
| 1145 |
+
"step": 360
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.3492957746478873,
|
| 1149 |
+
"forget_loss": 7.009506225585938e-05,
|
| 1150 |
+
"retain_loss": 5.6743621826171875e-05,
|
| 1151 |
+
"step": 360
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.3492957746478873,
|
| 1155 |
+
"forget_loss": 6.961822509765625e-05,
|
| 1156 |
+
"retain_loss": 4.00543212890625e-05,
|
| 1157 |
+
"step": 360
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.3868544600938968,
|
| 1161 |
+
"grad_norm": 0.012451171875,
|
| 1162 |
+
"learning_rate": 3.2101167315175102e-06,
|
| 1163 |
+
"loss": 0.00012967586517333984,
|
| 1164 |
+
"step": 370
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.3868544600938968,
|
| 1168 |
+
"forget_loss": 8.296966552734375e-05,
|
| 1169 |
+
"retain_loss": 5.435943603515625e-05,
|
| 1170 |
+
"step": 370
|
| 1171 |
+
},
|
| 1172 |
+
{
|
| 1173 |
+
"epoch": 1.3868544600938968,
|
| 1174 |
+
"forget_loss": 7.581710815429688e-05,
|
| 1175 |
+
"retain_loss": 3.647804260253906e-05,
|
| 1176 |
+
"step": 370
|
| 1177 |
+
},
|
| 1178 |
+
{
|
| 1179 |
+
"epoch": 1.3868544600938968,
|
| 1180 |
+
"forget_loss": 7.05718994140625e-05,
|
| 1181 |
+
"retain_loss": 3.075599670410156e-05,
|
| 1182 |
+
"step": 370
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 1.3868544600938968,
|
| 1186 |
+
"forget_loss": 7.104873657226562e-05,
|
| 1187 |
+
"retain_loss": 0.00010776519775390625,
|
| 1188 |
+
"step": 370
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 1.424413145539906,
|
| 1192 |
+
"grad_norm": 0.0162353515625,
|
| 1193 |
+
"learning_rate": 3.0155642023346305e-06,
|
| 1194 |
+
"loss": 0.0001344442367553711,
|
| 1195 |
+
"step": 380
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 1.424413145539906,
|
| 1199 |
+
"forget_loss": 7.62939453125e-05,
|
| 1200 |
+
"retain_loss": 5.793571472167969e-05,
|
| 1201 |
+
"step": 380
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.424413145539906,
|
| 1205 |
+
"forget_loss": 6.246566772460938e-05,
|
| 1206 |
+
"retain_loss": 3.838539123535156e-05,
|
| 1207 |
+
"step": 380
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 1.424413145539906,
|
| 1211 |
+
"forget_loss": 0.00010538101196289062,
|
| 1212 |
+
"retain_loss": 4.291534423828125e-05,
|
| 1213 |
+
"step": 380
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.424413145539906,
|
| 1217 |
+
"forget_loss": 7.009506225585938e-05,
|
| 1218 |
+
"retain_loss": 4.482269287109375e-05,
|
| 1219 |
+
"step": 380
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 1.4619718309859155,
|
| 1223 |
+
"grad_norm": 0.00567626953125,
|
| 1224 |
+
"learning_rate": 2.821011673151751e-06,
|
| 1225 |
+
"loss": 0.00012153387069702148,
|
| 1226 |
+
"step": 390
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.4619718309859155,
|
| 1230 |
+
"forget_loss": 8.440017700195312e-05,
|
| 1231 |
+
"retain_loss": 5.7697296142578125e-05,
|
| 1232 |
+
"step": 390
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 1.4619718309859155,
|
| 1236 |
+
"forget_loss": 7.390975952148438e-05,
|
| 1237 |
+
"retain_loss": 5.793571472167969e-05,
|
| 1238 |
+
"step": 390
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 1.4619718309859155,
|
| 1242 |
+
"forget_loss": 9.489059448242188e-05,
|
| 1243 |
+
"retain_loss": 3.647804260253906e-05,
|
| 1244 |
+
"step": 390
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 1.4619718309859155,
|
| 1248 |
+
"forget_loss": 9.059906005859375e-05,
|
| 1249 |
+
"retain_loss": 3.504753112792969e-05,
|
| 1250 |
+
"step": 390
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.4995305164319248,
|
| 1254 |
+
"grad_norm": 0.0308837890625,
|
| 1255 |
+
"learning_rate": 2.6264591439688717e-06,
|
| 1256 |
+
"loss": 0.00013158321380615234,
|
| 1257 |
+
"step": 400
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.4995305164319248,
|
| 1261 |
+
"forget_loss": 6.29425048828125e-05,
|
| 1262 |
+
"retain_loss": 4.4345855712890625e-05,
|
| 1263 |
+
"step": 400
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 1.4995305164319248,
|
| 1267 |
+
"forget_loss": 7.534027099609375e-05,
|
| 1268 |
+
"retain_loss": 4.5299530029296875e-05,
|
| 1269 |
+
"step": 400
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.4995305164319248,
|
| 1273 |
+
"forget_loss": 6.580352783203125e-05,
|
| 1274 |
+
"retain_loss": 4.673004150390625e-05,
|
| 1275 |
+
"step": 400
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 1.4995305164319248,
|
| 1279 |
+
"forget_loss": 6.961822509765625e-05,
|
| 1280 |
+
"retain_loss": 4.887580871582031e-05,
|
| 1281 |
+
"step": 400
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 1.5370892018779343,
|
| 1285 |
+
"grad_norm": 0.0159912109375,
|
| 1286 |
+
"learning_rate": 2.4319066147859924e-06,
|
| 1287 |
+
"loss": 0.00012365579605102539,
|
| 1288 |
+
"step": 410
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 1.5370892018779343,
|
| 1292 |
+
"forget_loss": 6.031990051269531e-05,
|
| 1293 |
+
"retain_loss": 3.266334533691406e-05,
|
| 1294 |
+
"step": 410
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 1.5370892018779343,
|
| 1298 |
+
"forget_loss": 8.296966552734375e-05,
|
| 1299 |
+
"retain_loss": 6.151199340820312e-05,
|
| 1300 |
+
"step": 410
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 1.5370892018779343,
|
| 1304 |
+
"forget_loss": 6.771087646484375e-05,
|
| 1305 |
+
"retain_loss": 4.673004150390625e-05,
|
| 1306 |
+
"step": 410
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.5370892018779343,
|
| 1310 |
+
"forget_loss": 6.532669067382812e-05,
|
| 1311 |
+
"retain_loss": 4.2438507080078125e-05,
|
| 1312 |
+
"step": 410
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 1.5746478873239438,
|
| 1316 |
+
"grad_norm": 0.00909423828125,
|
| 1317 |
+
"learning_rate": 2.237354085603113e-06,
|
| 1318 |
+
"loss": 0.0001282811164855957,
|
| 1319 |
+
"step": 420
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 1.5746478873239438,
|
| 1323 |
+
"forget_loss": 6.532669067382812e-05,
|
| 1324 |
+
"retain_loss": 3.0994415283203125e-05,
|
| 1325 |
+
"step": 420
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.5746478873239438,
|
| 1329 |
+
"forget_loss": 6.29425048828125e-05,
|
| 1330 |
+
"retain_loss": 3.337860107421875e-05,
|
| 1331 |
+
"step": 420
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.5746478873239438,
|
| 1335 |
+
"forget_loss": 7.581710815429688e-05,
|
| 1336 |
+
"retain_loss": 4.220008850097656e-05,
|
| 1337 |
+
"step": 420
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 1.5746478873239438,
|
| 1341 |
+
"forget_loss": 8.487701416015625e-05,
|
| 1342 |
+
"retain_loss": 3.4809112548828125e-05,
|
| 1343 |
+
"step": 420
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 1.612206572769953,
|
| 1347 |
+
"grad_norm": 0.007720947265625,
|
| 1348 |
+
"learning_rate": 2.0428015564202337e-06,
|
| 1349 |
+
"loss": 0.0001214742660522461,
|
| 1350 |
+
"step": 430
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 1.612206572769953,
|
| 1354 |
+
"forget_loss": 6.198883056640625e-05,
|
| 1355 |
+
"retain_loss": 4.839897155761719e-05,
|
| 1356 |
+
"step": 430
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 1.612206572769953,
|
| 1360 |
+
"forget_loss": 8.20159912109375e-05,
|
| 1361 |
+
"retain_loss": 5.364418029785156e-05,
|
| 1362 |
+
"step": 430
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.612206572769953,
|
| 1366 |
+
"forget_loss": 6.914138793945312e-05,
|
| 1367 |
+
"retain_loss": 5.626678466796875e-05,
|
| 1368 |
+
"step": 430
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 1.612206572769953,
|
| 1372 |
+
"forget_loss": 7.43865966796875e-05,
|
| 1373 |
+
"retain_loss": 3.170967102050781e-05,
|
| 1374 |
+
"step": 430
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.6497652582159623,
|
| 1378 |
+
"grad_norm": 0.0157470703125,
|
| 1379 |
+
"learning_rate": 1.8482490272373543e-06,
|
| 1380 |
+
"loss": 0.00011962652206420898,
|
| 1381 |
+
"step": 440
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.6497652582159623,
|
| 1385 |
+
"forget_loss": 7.486343383789062e-05,
|
| 1386 |
+
"retain_loss": 4.315376281738281e-05,
|
| 1387 |
+
"step": 440
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 1.6497652582159623,
|
| 1391 |
+
"forget_loss": 0.00011205673217773438,
|
| 1392 |
+
"retain_loss": 7.724761962890625e-05,
|
| 1393 |
+
"step": 440
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 1.6497652582159623,
|
| 1397 |
+
"forget_loss": 0.00010013580322265625,
|
| 1398 |
+
"retain_loss": 5.245208740234375e-05,
|
| 1399 |
+
"step": 440
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 1.6497652582159623,
|
| 1403 |
+
"forget_loss": 0.00010824203491210938,
|
| 1404 |
+
"retain_loss": 5.2928924560546875e-05,
|
| 1405 |
+
"step": 440
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 1.6873239436619718,
|
| 1409 |
+
"grad_norm": 0.017822265625,
|
| 1410 |
+
"learning_rate": 1.6536964980544747e-06,
|
| 1411 |
+
"loss": 0.000124359130859375,
|
| 1412 |
+
"step": 450
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 1.6873239436619718,
|
| 1416 |
+
"forget_loss": 9.202957153320312e-05,
|
| 1417 |
+
"retain_loss": 3.24249267578125e-05,
|
| 1418 |
+
"step": 450
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.6873239436619718,
|
| 1422 |
+
"forget_loss": 6.151199340820312e-05,
|
| 1423 |
+
"retain_loss": 4.267692565917969e-05,
|
| 1424 |
+
"step": 450
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.6873239436619718,
|
| 1428 |
+
"forget_loss": 6.437301635742188e-05,
|
| 1429 |
+
"retain_loss": 5.245208740234375e-05,
|
| 1430 |
+
"step": 450
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.6873239436619718,
|
| 1434 |
+
"forget_loss": 5.698204040527344e-05,
|
| 1435 |
+
"retain_loss": 6.341934204101562e-05,
|
| 1436 |
+
"step": 450
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 1.7248826291079813,
|
| 1440 |
+
"grad_norm": 0.0067138671875,
|
| 1441 |
+
"learning_rate": 1.4591439688715956e-06,
|
| 1442 |
+
"loss": 0.00012021064758300781,
|
| 1443 |
+
"step": 460
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 1.7248826291079813,
|
| 1447 |
+
"forget_loss": 6.914138793945312e-05,
|
| 1448 |
+
"retain_loss": 5.125999450683594e-05,
|
| 1449 |
+
"step": 460
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 1.7248826291079813,
|
| 1453 |
+
"forget_loss": 7.772445678710938e-05,
|
| 1454 |
+
"retain_loss": 3.647804260253906e-05,
|
| 1455 |
+
"step": 460
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.7248826291079813,
|
| 1459 |
+
"forget_loss": 7.200241088867188e-05,
|
| 1460 |
+
"retain_loss": 3.266334533691406e-05,
|
| 1461 |
+
"step": 460
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 1.7248826291079813,
|
| 1465 |
+
"forget_loss": 8.678436279296875e-05,
|
| 1466 |
+
"retain_loss": 7.581710815429688e-05,
|
| 1467 |
+
"step": 460
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.7624413145539906,
|
| 1471 |
+
"grad_norm": 0.01202392578125,
|
| 1472 |
+
"learning_rate": 1.264591439688716e-06,
|
| 1473 |
+
"loss": 0.0001221299171447754,
|
| 1474 |
+
"step": 470
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.7624413145539906,
|
| 1478 |
+
"forget_loss": 8.678436279296875e-05,
|
| 1479 |
+
"retain_loss": 4.363059997558594e-05,
|
| 1480 |
+
"step": 470
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.7624413145539906,
|
| 1484 |
+
"forget_loss": 6.723403930664062e-05,
|
| 1485 |
+
"retain_loss": 3.337860107421875e-05,
|
| 1486 |
+
"step": 470
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 1.7624413145539906,
|
| 1490 |
+
"forget_loss": 7.343292236328125e-05,
|
| 1491 |
+
"retain_loss": 8.249282836914062e-05,
|
| 1492 |
+
"step": 470
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 1.7624413145539906,
|
| 1496 |
+
"forget_loss": 7.581710815429688e-05,
|
| 1497 |
+
"retain_loss": 2.682209014892578e-05,
|
| 1498 |
+
"step": 470
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 1.8,
|
| 1502 |
+
"grad_norm": 0.00830078125,
|
| 1503 |
+
"learning_rate": 1.0700389105058367e-06,
|
| 1504 |
+
"loss": 0.00011826753616333008,
|
| 1505 |
+
"step": 480
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1.8,
|
| 1509 |
+
"forget_loss": 7.390975952148438e-05,
|
| 1510 |
+
"retain_loss": 3.3855438232421875e-05,
|
| 1511 |
+
"step": 480
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 1.8,
|
| 1515 |
+
"forget_loss": 6.723403930664062e-05,
|
| 1516 |
+
"retain_loss": 9.059906005859375e-05,
|
| 1517 |
+
"step": 480
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 1.8,
|
| 1521 |
+
"forget_loss": 8.0108642578125e-05,
|
| 1522 |
+
"retain_loss": 4.7206878662109375e-05,
|
| 1523 |
+
"step": 480
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 1.8,
|
| 1527 |
+
"forget_loss": 6.151199340820312e-05,
|
| 1528 |
+
"retain_loss": 4.76837158203125e-05,
|
| 1529 |
+
"step": 480
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.8375586854460093,
|
| 1533 |
+
"grad_norm": 0.0045166015625,
|
| 1534 |
+
"learning_rate": 8.754863813229572e-07,
|
| 1535 |
+
"loss": 0.00012172460556030274,
|
| 1536 |
+
"step": 490
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 1.8375586854460093,
|
| 1540 |
+
"forget_loss": 7.05718994140625e-05,
|
| 1541 |
+
"retain_loss": 4.315376281738281e-05,
|
| 1542 |
+
"step": 490
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 1.8375586854460093,
|
| 1546 |
+
"forget_loss": 5.793571472167969e-05,
|
| 1547 |
+
"retain_loss": 3.5762786865234375e-05,
|
| 1548 |
+
"step": 490
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 1.8375586854460093,
|
| 1552 |
+
"forget_loss": 6.008148193359375e-05,
|
| 1553 |
+
"retain_loss": 3.4809112548828125e-05,
|
| 1554 |
+
"step": 490
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 1.8375586854460093,
|
| 1558 |
+
"forget_loss": 7.82012939453125e-05,
|
| 1559 |
+
"retain_loss": 3.838539123535156e-05,
|
| 1560 |
+
"step": 490
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 1.8751173708920188,
|
| 1564 |
+
"grad_norm": 0.01422119140625,
|
| 1565 |
+
"learning_rate": 6.809338521400779e-07,
|
| 1566 |
+
"loss": 0.00011525154113769531,
|
| 1567 |
+
"step": 500
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 1.8751173708920188,
|
| 1571 |
+
"forget_loss": 6.818771362304688e-05,
|
| 1572 |
+
"retain_loss": 3.838539123535156e-05,
|
| 1573 |
+
"step": 500
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 1.8751173708920188,
|
| 1577 |
+
"forget_loss": 9.489059448242188e-05,
|
| 1578 |
+
"retain_loss": 8.58306884765625e-05,
|
| 1579 |
+
"step": 500
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 1.8751173708920188,
|
| 1583 |
+
"forget_loss": 9.298324584960938e-05,
|
| 1584 |
+
"retain_loss": 4.4345855712890625e-05,
|
| 1585 |
+
"step": 500
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.8751173708920188,
|
| 1589 |
+
"forget_loss": 6.151199340820312e-05,
|
| 1590 |
+
"retain_loss": 6.723403930664062e-05,
|
| 1591 |
+
"step": 500
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 1.9126760563380283,
|
| 1595 |
+
"grad_norm": 0.01080322265625,
|
| 1596 |
+
"learning_rate": 4.863813229571985e-07,
|
| 1597 |
+
"loss": 0.00011955499649047851,
|
| 1598 |
+
"step": 510
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 1.9126760563380283,
|
| 1602 |
+
"forget_loss": 7.62939453125e-05,
|
| 1603 |
+
"retain_loss": 4.2438507080078125e-05,
|
| 1604 |
+
"step": 510
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 1.9126760563380283,
|
| 1608 |
+
"forget_loss": 8.630752563476562e-05,
|
| 1609 |
+
"retain_loss": 4.863739013671875e-05,
|
| 1610 |
+
"step": 510
|
| 1611 |
+
},
|
| 1612 |
+
{
|
| 1613 |
+
"epoch": 1.9126760563380283,
|
| 1614 |
+
"forget_loss": 8.96453857421875e-05,
|
| 1615 |
+
"retain_loss": 5.1021575927734375e-05,
|
| 1616 |
+
"step": 510
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 1.9126760563380283,
|
| 1620 |
+
"forget_loss": 7.677078247070312e-05,
|
| 1621 |
+
"retain_loss": 4.076957702636719e-05,
|
| 1622 |
+
"step": 510
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 1.9502347417840376,
|
| 1626 |
+
"grad_norm": 0.006256103515625,
|
| 1627 |
+
"learning_rate": 2.918287937743191e-07,
|
| 1628 |
+
"loss": 0.00011934041976928711,
|
| 1629 |
+
"step": 520
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 1.9502347417840376,
|
| 1633 |
+
"forget_loss": 7.152557373046875e-05,
|
| 1634 |
+
"retain_loss": 5.8650970458984375e-05,
|
| 1635 |
+
"step": 520
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.9502347417840376,
|
| 1639 |
+
"forget_loss": 7.343292236328125e-05,
|
| 1640 |
+
"retain_loss": 4.553794860839844e-05,
|
| 1641 |
+
"step": 520
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 1.9502347417840376,
|
| 1645 |
+
"forget_loss": 6.771087646484375e-05,
|
| 1646 |
+
"retain_loss": 3.647804260253906e-05,
|
| 1647 |
+
"step": 520
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 1.9502347417840376,
|
| 1651 |
+
"forget_loss": 6.29425048828125e-05,
|
| 1652 |
+
"retain_loss": 3.361701965332031e-05,
|
| 1653 |
+
"step": 520
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 1.9877934272300468,
|
| 1657 |
+
"grad_norm": 0.01611328125,
|
| 1658 |
+
"learning_rate": 9.727626459143969e-08,
|
| 1659 |
+
"loss": 0.00012475252151489258,
|
| 1660 |
+
"step": 530
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 1.9877934272300468,
|
| 1664 |
+
"forget_loss": 7.104873657226562e-05,
|
| 1665 |
+
"retain_loss": 5.745887756347656e-05,
|
| 1666 |
+
"step": 530
|
| 1667 |
+
},
|
| 1668 |
+
{
|
| 1669 |
+
"epoch": 1.9877934272300468,
|
| 1670 |
+
"forget_loss": 8.296966552734375e-05,
|
| 1671 |
+
"retain_loss": 3.6716461181640625e-05,
|
| 1672 |
+
"step": 530
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 1.9877934272300468,
|
| 1676 |
+
"forget_loss": 6.818771362304688e-05,
|
| 1677 |
+
"retain_loss": 5.1021575927734375e-05,
|
| 1678 |
+
"step": 530
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 1.9877934272300468,
|
| 1682 |
+
"forget_loss": 6.437301635742188e-05,
|
| 1683 |
+
"retain_loss": 4.601478576660156e-05,
|
| 1684 |
+
"step": 530
|
| 1685 |
+
}
|
| 1686 |
+
],
|
| 1687 |
+
"logging_steps": 10,
|
| 1688 |
+
"max_steps": 534,
|
| 1689 |
+
"num_input_tokens_seen": 0,
|
| 1690 |
+
"num_train_epochs": 2,
|
| 1691 |
+
"save_steps": 200,
|
| 1692 |
+
"stateful_callbacks": {
|
| 1693 |
+
"TrainerControl": {
|
| 1694 |
+
"args": {
|
| 1695 |
+
"should_epoch_stop": false,
|
| 1696 |
+
"should_evaluate": false,
|
| 1697 |
+
"should_log": false,
|
| 1698 |
+
"should_save": true,
|
| 1699 |
+
"should_training_stop": true
|
| 1700 |
+
},
|
| 1701 |
+
"attributes": {}
|
| 1702 |
+
}
|
| 1703 |
+
},
|
| 1704 |
+
"total_flos": 0.0,
|
| 1705 |
+
"train_batch_size": 8,
|
| 1706 |
+
"trial_name": null,
|
| 1707 |
+
"trial_params": null
|
| 1708 |
+
}
|
checkpoint-534/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cde3986b5d2902c3143ac3a5311b9443a600348ae4fbcaab8a20b0760c034ff1
|
| 3 |
+
size 5265
|
config.json
ADDED
|
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"LlamaForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 128000,
|
| 8 |
+
"dtype": "bfloat16",
|
| 9 |
+
"eos_token_id": 128001,
|
| 10 |
+
"head_dim": 128,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 4096,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 14336,
|
| 15 |
+
"max_position_embeddings": 131072,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 32,
|
| 19 |
+
"num_hidden_layers": 32,
|
| 20 |
+
"num_key_value_heads": 8,
|
| 21 |
+
"pad_token_id": 128001,
|
| 22 |
+
"pretraining_tp": 1,
|
| 23 |
+
"rms_norm_eps": 1e-05,
|
| 24 |
+
"rope_parameters": {
|
| 25 |
+
"factor": 8.0,
|
| 26 |
+
"high_freq_factor": 4.0,
|
| 27 |
+
"low_freq_factor": 1.0,
|
| 28 |
+
"original_max_position_embeddings": 8192,
|
| 29 |
+
"rope_theta": 500000.0,
|
| 30 |
+
"rope_type": "llama3"
|
| 31 |
+
},
|
| 32 |
+
"tie_word_embeddings": false,
|
| 33 |
+
"transformers_version": "5.6.0",
|
| 34 |
+
"use_cache": false,
|
| 35 |
+
"vocab_size": 128256
|
| 36 |
+
}
|
config.yaml
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
model_name: meta-llama/Llama-3.1-8B
|
| 2 |
+
forget_path: unlearning-cleanslate/fsid-curated-llama-8b:forget:baseline
|
| 3 |
+
retain_path: unlearning-cleanslate/fsid-curated-llama-8b:retain:baseline
|
| 4 |
+
output_dir: /runs
|
| 5 |
+
objective: rmu
|
| 6 |
+
num_epochs: 2
|
| 7 |
+
batch_size: 8
|
| 8 |
+
gradient_accum_steps: 4
|
| 9 |
+
lr: 1.0e-05
|
| 10 |
+
weight_decay: 0.0
|
| 11 |
+
warmup_steps: 20
|
| 12 |
+
max_grad_norm: 1.0
|
| 13 |
+
forget_weight: 1.0
|
| 14 |
+
retain_weight: 1.0
|
| 15 |
+
retain_loss_type: embed_diff
|
| 16 |
+
npo_beta: 0.1
|
| 17 |
+
simnpo_delta: 0.0
|
| 18 |
+
rmu_steering_coeff: 2.0
|
| 19 |
+
rmu_module_regex: model\.layers\.7
|
| 20 |
+
rmu_control_vec_seed: 0
|
| 21 |
+
undial_beta: 10.0
|
| 22 |
+
save_every: 200
|
| 23 |
+
fsdp_sharding_strategy: full_shard
|
| 24 |
+
dtype: bfloat16
|
| 25 |
+
seed: 42
|
| 26 |
+
num_workers: 0
|
| 27 |
+
log_every: 10
|
| 28 |
+
gradient_checkpointing: false
|
| 29 |
+
torch_compile: false
|
| 30 |
+
push_to_hub: true
|
| 31 |
+
hub_repo_id: unlearning-cleanslate/llama-3_1-8b-rmu-baseline
|
| 32 |
+
hub_private: false
|
| 33 |
+
optim: adamw_torch
|
| 34 |
+
_run_name: rmu_Llama-3.1-8B_20260427_203908
|
| 35 |
+
_run_dir: /runs/rmu_Llama-3.1-8B_20260427_203908
|
generation_config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 128000,
|
| 4 |
+
"do_sample": true,
|
| 5 |
+
"eos_token_id": [
|
| 6 |
+
128001
|
| 7 |
+
],
|
| 8 |
+
"pad_token_id": 128001,
|
| 9 |
+
"temperature": 0.6,
|
| 10 |
+
"top_p": 0.9,
|
| 11 |
+
"transformers_version": "5.6.0"
|
| 12 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f4b6fddb151d0b4bd2f8c5129e10e07f980a758209406b9308bd5b85a60af1f6
|
| 3 |
+
size 16060556616
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b
|
| 3 |
+
size 17209920
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "tokenizers",
|
| 3 |
+
"bos_token": "<|begin_of_text|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
+
"eos_token": "<|end_of_text|>",
|
| 6 |
+
"is_local": false,
|
| 7 |
+
"local_files_only": false,
|
| 8 |
+
"model_input_names": [
|
| 9 |
+
"input_ids",
|
| 10 |
+
"attention_mask"
|
| 11 |
+
],
|
| 12 |
+
"model_max_length": 131072,
|
| 13 |
+
"pad_token": "<|end_of_text|>",
|
| 14 |
+
"tokenizer_class": "TokenizersBackend"
|
| 15 |
+
}
|
train.log
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
2026-04-27 20:39:32,701 INFO cleanslate.fsid.training.trainer: [RANK 0] FSID train: model=meta-llama/Llama-3.1-8B objective=rmu run_dir=/runs/rmu_Llama-3.1-8B_20260427_203908
|
| 2 |
+
2026-04-27 20:39:32,703 INFO cleanslate.fsid.training.trainer: [RANK 0] Creating frozen reference model for rmu (dtype=torch.bfloat16)
|
| 3 |
+
2026-04-27 20:39:37,857 INFO cleanslate.fsid.training.trainer: [RANK 0] Starting training: objective=rmu epochs=2
|
| 4 |
+
2026-04-27 20:39:38,020 INFO cleanslate.fsid.training.trainer: [RANK 0] RMU control_vec init: dim=4096 seed=0 rank=0 world_size=1 per_rank_norm=2.000000
|
| 5 |
+
2026-04-27 20:50:17,489 INFO cleanslate.fsid.training.trainer: [RANK 0] Saved checkpoint to /runs/rmu_Llama-3.1-8B_20260427_203908
|
| 6 |
+
2026-04-27 20:50:17,498 INFO cleanslate.fsid.training.trainer: [RANK 0] Training complete (586.7s)
|
| 7 |
+
2026-04-27 20:50:17,901 INFO cleanslate.fsid.training.trainer: [RANK 0] Pushing checkpoint dir /runs/rmu_Llama-3.1-8B_20260427_203908 to unlearning-cleanslate/llama-3_1-8b-rmu-baseline
|
train_summary.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"run_name": "rmu_Llama-3.1-8B_20260427_203908",
|
| 3 |
+
"run_dir": "/runs/rmu_Llama-3.1-8B_20260427_203908",
|
| 4 |
+
"config": {
|
| 5 |
+
"model_name": "meta-llama/Llama-3.1-8B",
|
| 6 |
+
"forget_path": "unlearning-cleanslate/fsid-curated-llama-8b:forget:baseline",
|
| 7 |
+
"retain_path": "unlearning-cleanslate/fsid-curated-llama-8b:retain:baseline",
|
| 8 |
+
"output_dir": "/runs",
|
| 9 |
+
"objective": "rmu",
|
| 10 |
+
"num_epochs": 2,
|
| 11 |
+
"batch_size": 8,
|
| 12 |
+
"gradient_accum_steps": 4,
|
| 13 |
+
"lr": 1e-05,
|
| 14 |
+
"weight_decay": 0.0,
|
| 15 |
+
"warmup_steps": 20,
|
| 16 |
+
"max_grad_norm": 1.0,
|
| 17 |
+
"forget_weight": 1.0,
|
| 18 |
+
"retain_weight": 1.0,
|
| 19 |
+
"retain_loss_type": "embed_diff",
|
| 20 |
+
"npo_beta": 0.1,
|
| 21 |
+
"simnpo_delta": 0.0,
|
| 22 |
+
"rmu_steering_coeff": 2.0,
|
| 23 |
+
"rmu_module_regex": "model\\.layers\\.7",
|
| 24 |
+
"rmu_control_vec_seed": 0,
|
| 25 |
+
"undial_beta": 10.0,
|
| 26 |
+
"save_every": 200,
|
| 27 |
+
"fsdp_sharding_strategy": "full_shard",
|
| 28 |
+
"dtype": "bfloat16",
|
| 29 |
+
"seed": 42,
|
| 30 |
+
"num_workers": 0,
|
| 31 |
+
"log_every": 10,
|
| 32 |
+
"gradient_checkpointing": false,
|
| 33 |
+
"torch_compile": false,
|
| 34 |
+
"push_to_hub": true,
|
| 35 |
+
"hub_repo_id": "unlearning-cleanslate/llama-3_1-8b-rmu-baseline",
|
| 36 |
+
"hub_private": false,
|
| 37 |
+
"optim": "adamw_torch"
|
| 38 |
+
},
|
| 39 |
+
"num_forget_rows": 8513,
|
| 40 |
+
"num_forget_pairs": 8513,
|
| 41 |
+
"num_retain_rows": 1646,
|
| 42 |
+
"num_retain_sequences": 1646,
|
| 43 |
+
"num_trainable_params": 8030261248,
|
| 44 |
+
"num_total_params": 8030261248,
|
| 45 |
+
"run_seconds": 586.7230443954468,
|
| 46 |
+
"checkpoint_dir": "/runs/rmu_Llama-3.1-8B_20260427_203908",
|
| 47 |
+
"is_main_process": true
|
| 48 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cde3986b5d2902c3143ac3a5311b9443a600348ae4fbcaab8a20b0760c034ff1
|
| 3 |
+
size 5265
|