Upload folder using huggingface_hub
#1
by
sameepv21
- opened
- README.md +9 -0
- adapter_config.json +29 -0
- adapter_model.bin +3 -0
- config.json +56 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +439 -0
README.md
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
---
|
| 4 |
+
## Training procedure
|
| 5 |
+
|
| 6 |
+
### Framework versions
|
| 7 |
+
|
| 8 |
+
|
| 9 |
+
- PEFT 0.5.0
|
adapter_config.json
ADDED
|
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "DAMO-NLP-SG/VideoLLaMA3-7B",
|
| 4 |
+
"bias": "none",
|
| 5 |
+
"fan_in_fan_out": false,
|
| 6 |
+
"inference_mode": true,
|
| 7 |
+
"init_lora_weights": true,
|
| 8 |
+
"layers_pattern": null,
|
| 9 |
+
"layers_to_transform": null,
|
| 10 |
+
"lora_alpha": 256,
|
| 11 |
+
"lora_dropout": 0.05,
|
| 12 |
+
"modules_to_save": null,
|
| 13 |
+
"peft_type": "LORA",
|
| 14 |
+
"r": 128,
|
| 15 |
+
"revision": null,
|
| 16 |
+
"target_modules": [
|
| 17 |
+
"out_proj",
|
| 18 |
+
"up_proj",
|
| 19 |
+
"q_proj",
|
| 20 |
+
"down_proj",
|
| 21 |
+
"fc2",
|
| 22 |
+
"v_proj",
|
| 23 |
+
"fc1",
|
| 24 |
+
"gate_proj",
|
| 25 |
+
"o_proj",
|
| 26 |
+
"k_proj"
|
| 27 |
+
],
|
| 28 |
+
"task_type": "CAUSAL_LM"
|
| 29 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e982e2b4e9e8c73e9e973b3285df8bd845569bc86a8acc39791fe8c4d32d1c5e
|
| 3 |
+
size 785303482
|
config.json
ADDED
|
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"X": [
|
| 3 |
+
"Image",
|
| 4 |
+
"Video"
|
| 5 |
+
],
|
| 6 |
+
"_attn_implementation_autoset": true,
|
| 7 |
+
"_name_or_path": "DAMO-NLP-SG/VideoLLaMA3-7B",
|
| 8 |
+
"architectures": [
|
| 9 |
+
"Videollama3Qwen2ForCausalLM"
|
| 10 |
+
],
|
| 11 |
+
"attention_dropout": 0.0,
|
| 12 |
+
"auto_map": {
|
| 13 |
+
"AutoConfig": "DAMO-NLP-SG/VideoLLaMA3-7B--configuration_videollama3.Videollama3Qwen2Config",
|
| 14 |
+
"AutoModelForCausalLM": "DAMO-NLP-SG/VideoLLaMA3-7B--modeling_videollama3.Videollama3Qwen2ForCausalLM"
|
| 15 |
+
},
|
| 16 |
+
"bos_token_id": 151643,
|
| 17 |
+
"eos_token_id": 151645,
|
| 18 |
+
"freeze_mm_mlp_adapter": true,
|
| 19 |
+
"hidden_act": "silu",
|
| 20 |
+
"hidden_size": 3584,
|
| 21 |
+
"image_aspect_ratio": "pad",
|
| 22 |
+
"image_grid_pinpoints": null,
|
| 23 |
+
"image_token_index": 151665,
|
| 24 |
+
"initializer_range": 0.02,
|
| 25 |
+
"intermediate_size": 18944,
|
| 26 |
+
"max_position_embeddings": 32768,
|
| 27 |
+
"max_window_layers": 28,
|
| 28 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 29 |
+
"mm_use_x_patch_token": false,
|
| 30 |
+
"mm_use_x_start_end": false,
|
| 31 |
+
"model_type": "videollama3_qwen2",
|
| 32 |
+
"num_attention_heads": 28,
|
| 33 |
+
"num_hidden_layers": 28,
|
| 34 |
+
"num_key_value_heads": 4,
|
| 35 |
+
"rms_norm_eps": 1e-06,
|
| 36 |
+
"rope_scaling": null,
|
| 37 |
+
"rope_theta": 1000000.0,
|
| 38 |
+
"sliding_window": null,
|
| 39 |
+
"tie_word_embeddings": false,
|
| 40 |
+
"torch_dtype": "bfloat16",
|
| 41 |
+
"transformers_version": "4.46.3",
|
| 42 |
+
"tune_mm_mlp_adapter": false,
|
| 43 |
+
"use_cache": true,
|
| 44 |
+
"use_sliding_window": false,
|
| 45 |
+
"use_token_compression": false,
|
| 46 |
+
"vision_encoder": null,
|
| 47 |
+
"vision_encoder_config": {
|
| 48 |
+
"hidden_size": 1152,
|
| 49 |
+
"intermediate_size": 4304,
|
| 50 |
+
"model_type": "videollama3_vision_encoder",
|
| 51 |
+
"num_attention_heads": 16,
|
| 52 |
+
"num_hidden_layers": 27,
|
| 53 |
+
"patch_size": 14
|
| 54 |
+
},
|
| 55 |
+
"vocab_size": 152064
|
| 56 |
+
}
|
non_lora_trainables.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60fb82c3660319e6d0b239950b20c28181e97f1ade117dc0660b40e2ad94a89b
|
| 3 |
+
size 912
|
trainer_state.json
ADDED
|
@@ -0,0 +1,439 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9927360774818403,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 618,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.12,
|
| 13 |
+
"learning_rate": 2.0161290322580642e-07,
|
| 14 |
+
"logps/chosen": -155.47128295898438,
|
| 15 |
+
"logps/rejected": -104.8045883178711,
|
| 16 |
+
"loss": 0.7536,
|
| 17 |
+
"losses/dpo": 0.7361882925033569,
|
| 18 |
+
"losses/sft": 1.3545048236846924,
|
| 19 |
+
"losses/total": 0.7361882925033569,
|
| 20 |
+
"ref_logps/chosen": -155.36932373046875,
|
| 21 |
+
"ref_logps/rejected": -104.93973541259766,
|
| 22 |
+
"rewards/accuracies": 0.48450005054473877,
|
| 23 |
+
"rewards/chosen": -0.010196244344115257,
|
| 24 |
+
"rewards/margins": -0.023710083216428757,
|
| 25 |
+
"rewards/rejected": 0.0135138388723135,
|
| 26 |
+
"step": 25
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.24,
|
| 30 |
+
"learning_rate": 4.0322580645161285e-07,
|
| 31 |
+
"logps/chosen": -153.7573699951172,
|
| 32 |
+
"logps/rejected": -107.0270004272461,
|
| 33 |
+
"loss": 0.7401,
|
| 34 |
+
"losses/dpo": 0.7450882792472839,
|
| 35 |
+
"losses/sft": 1.3266704082489014,
|
| 36 |
+
"losses/total": 0.7450882792472839,
|
| 37 |
+
"ref_logps/chosen": -153.8381805419922,
|
| 38 |
+
"ref_logps/rejected": -107.06134033203125,
|
| 39 |
+
"rewards/accuracies": 0.502500057220459,
|
| 40 |
+
"rewards/chosen": 0.00808356050401926,
|
| 41 |
+
"rewards/margins": 0.004649868700653315,
|
| 42 |
+
"rewards/rejected": 0.003433691570535302,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.36,
|
| 47 |
+
"learning_rate": 4.883093525179856e-07,
|
| 48 |
+
"logps/chosen": -151.94009399414062,
|
| 49 |
+
"logps/rejected": -105.40670013427734,
|
| 50 |
+
"loss": 0.7213,
|
| 51 |
+
"losses/dpo": 0.7041549682617188,
|
| 52 |
+
"losses/sft": 1.3607875108718872,
|
| 53 |
+
"losses/total": 0.7041549682617188,
|
| 54 |
+
"ref_logps/chosen": -152.60272216796875,
|
| 55 |
+
"ref_logps/rejected": -105.59368896484375,
|
| 56 |
+
"rewards/accuracies": 0.5264999866485596,
|
| 57 |
+
"rewards/chosen": 0.06626255810260773,
|
| 58 |
+
"rewards/margins": 0.04756266996264458,
|
| 59 |
+
"rewards/rejected": 0.01869989186525345,
|
| 60 |
+
"step": 75
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.48,
|
| 64 |
+
"learning_rate": 4.6582733812949637e-07,
|
| 65 |
+
"logps/chosen": -149.70230102539062,
|
| 66 |
+
"logps/rejected": -106.14666748046875,
|
| 67 |
+
"loss": 0.6625,
|
| 68 |
+
"losses/dpo": 0.6611780524253845,
|
| 69 |
+
"losses/sft": 1.341475009918213,
|
| 70 |
+
"losses/total": 0.6611780524253845,
|
| 71 |
+
"ref_logps/chosen": -152.02659606933594,
|
| 72 |
+
"ref_logps/rejected": -106.69593811035156,
|
| 73 |
+
"rewards/accuracies": 0.621999979019165,
|
| 74 |
+
"rewards/chosen": 0.2324293702840805,
|
| 75 |
+
"rewards/margins": 0.17750366032123566,
|
| 76 |
+
"rewards/rejected": 0.05492572858929634,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.61,
|
| 81 |
+
"learning_rate": 4.4334532374100717e-07,
|
| 82 |
+
"logps/chosen": -148.40090942382812,
|
| 83 |
+
"logps/rejected": -104.86791229248047,
|
| 84 |
+
"loss": 0.597,
|
| 85 |
+
"losses/dpo": 0.5830589532852173,
|
| 86 |
+
"losses/sft": 1.2935805320739746,
|
| 87 |
+
"losses/total": 0.5830589532852173,
|
| 88 |
+
"ref_logps/chosen": -152.3332977294922,
|
| 89 |
+
"ref_logps/rejected": -105.52149200439453,
|
| 90 |
+
"rewards/accuracies": 0.7135000824928284,
|
| 91 |
+
"rewards/chosen": 0.39323729276657104,
|
| 92 |
+
"rewards/margins": 0.32787904143333435,
|
| 93 |
+
"rewards/rejected": 0.0653582364320755,
|
| 94 |
+
"step": 125
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.73,
|
| 98 |
+
"learning_rate": 4.2086330935251797e-07,
|
| 99 |
+
"logps/chosen": -147.95791625976562,
|
| 100 |
+
"logps/rejected": -106.17552947998047,
|
| 101 |
+
"loss": 0.5515,
|
| 102 |
+
"losses/dpo": 0.5297109484672546,
|
| 103 |
+
"losses/sft": 1.289250135421753,
|
| 104 |
+
"losses/total": 0.5297109484672546,
|
| 105 |
+
"ref_logps/chosen": -153.35595703125,
|
| 106 |
+
"ref_logps/rejected": -106.98403930664062,
|
| 107 |
+
"rewards/accuracies": 0.7540000677108765,
|
| 108 |
+
"rewards/chosen": 0.5398061871528625,
|
| 109 |
+
"rewards/margins": 0.45895519852638245,
|
| 110 |
+
"rewards/rejected": 0.0808510109782219,
|
| 111 |
+
"step": 150
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.85,
|
| 115 |
+
"learning_rate": 3.9838129496402877e-07,
|
| 116 |
+
"logps/chosen": -147.28538513183594,
|
| 117 |
+
"logps/rejected": -106.73645782470703,
|
| 118 |
+
"loss": 0.5171,
|
| 119 |
+
"losses/dpo": 0.526803731918335,
|
| 120 |
+
"losses/sft": 1.2763028144836426,
|
| 121 |
+
"losses/total": 0.526803731918335,
|
| 122 |
+
"ref_logps/chosen": -153.7596893310547,
|
| 123 |
+
"ref_logps/rejected": -107.54974365234375,
|
| 124 |
+
"rewards/accuracies": 0.7815000414848328,
|
| 125 |
+
"rewards/chosen": 0.6474303603172302,
|
| 126 |
+
"rewards/margins": 0.5661011934280396,
|
| 127 |
+
"rewards/rejected": 0.0813291072845459,
|
| 128 |
+
"step": 175
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.97,
|
| 132 |
+
"learning_rate": 3.7589928057553957e-07,
|
| 133 |
+
"logps/chosen": -146.67481994628906,
|
| 134 |
+
"logps/rejected": -104.6893539428711,
|
| 135 |
+
"loss": 0.4544,
|
| 136 |
+
"losses/dpo": 0.4492318332195282,
|
| 137 |
+
"losses/sft": 1.284212350845337,
|
| 138 |
+
"losses/total": 0.4492318332195282,
|
| 139 |
+
"ref_logps/chosen": -154.76519775390625,
|
| 140 |
+
"ref_logps/rejected": -105.09065246582031,
|
| 141 |
+
"rewards/accuracies": 0.825499951839447,
|
| 142 |
+
"rewards/chosen": 0.809037446975708,
|
| 143 |
+
"rewards/margins": 0.7689078450202942,
|
| 144 |
+
"rewards/rejected": 0.040129706263542175,
|
| 145 |
+
"step": 200
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 1.09,
|
| 149 |
+
"learning_rate": 3.5341726618705037e-07,
|
| 150 |
+
"logps/chosen": -144.06991577148438,
|
| 151 |
+
"logps/rejected": -105.81298828125,
|
| 152 |
+
"loss": 0.4189,
|
| 153 |
+
"losses/dpo": 0.42898857593536377,
|
| 154 |
+
"losses/sft": 1.2516932487487793,
|
| 155 |
+
"losses/total": 0.42898857593536377,
|
| 156 |
+
"ref_logps/chosen": -153.25570678710938,
|
| 157 |
+
"ref_logps/rejected": -106.06501007080078,
|
| 158 |
+
"rewards/accuracies": 0.8469999432563782,
|
| 159 |
+
"rewards/chosen": 0.9185801148414612,
|
| 160 |
+
"rewards/margins": 0.8933786749839783,
|
| 161 |
+
"rewards/rejected": 0.02520136535167694,
|
| 162 |
+
"step": 225
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 1.21,
|
| 166 |
+
"learning_rate": 3.309352517985611e-07,
|
| 167 |
+
"logps/chosen": -144.94163513183594,
|
| 168 |
+
"logps/rejected": -106.42683410644531,
|
| 169 |
+
"loss": 0.3982,
|
| 170 |
+
"losses/dpo": 0.41044384241104126,
|
| 171 |
+
"losses/sft": 1.2750244140625,
|
| 172 |
+
"losses/total": 0.41044384241104126,
|
| 173 |
+
"ref_logps/chosen": -154.86021423339844,
|
| 174 |
+
"ref_logps/rejected": -106.34297180175781,
|
| 175 |
+
"rewards/accuracies": 0.8530000448226929,
|
| 176 |
+
"rewards/chosen": 0.9918593764305115,
|
| 177 |
+
"rewards/margins": 1.0002468824386597,
|
| 178 |
+
"rewards/rejected": -0.008387637324631214,
|
| 179 |
+
"step": 250
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.33,
|
| 183 |
+
"learning_rate": 3.0845323741007197e-07,
|
| 184 |
+
"logps/chosen": -144.10977172851562,
|
| 185 |
+
"logps/rejected": -105.52392578125,
|
| 186 |
+
"loss": 0.3652,
|
| 187 |
+
"losses/dpo": 0.3518593907356262,
|
| 188 |
+
"losses/sft": 1.2418614625930786,
|
| 189 |
+
"losses/total": 0.3518593907356262,
|
| 190 |
+
"ref_logps/chosen": -154.4581298828125,
|
| 191 |
+
"ref_logps/rejected": -104.62450408935547,
|
| 192 |
+
"rewards/accuracies": 0.8764999508857727,
|
| 193 |
+
"rewards/chosen": 1.0348376035690308,
|
| 194 |
+
"rewards/margins": 1.1247813701629639,
|
| 195 |
+
"rewards/rejected": -0.08994373679161072,
|
| 196 |
+
"step": 275
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.45,
|
| 200 |
+
"learning_rate": 2.859712230215827e-07,
|
| 201 |
+
"logps/chosen": -142.1906280517578,
|
| 202 |
+
"logps/rejected": -108.99441528320312,
|
| 203 |
+
"loss": 0.3624,
|
| 204 |
+
"losses/dpo": 0.35955867171287537,
|
| 205 |
+
"losses/sft": 1.2526122331619263,
|
| 206 |
+
"losses/total": 0.35955867171287537,
|
| 207 |
+
"ref_logps/chosen": -152.7955322265625,
|
| 208 |
+
"ref_logps/rejected": -107.89449310302734,
|
| 209 |
+
"rewards/accuracies": 0.8740000128746033,
|
| 210 |
+
"rewards/chosen": 1.0604920387268066,
|
| 211 |
+
"rewards/margins": 1.170485496520996,
|
| 212 |
+
"rewards/rejected": -0.10999350249767303,
|
| 213 |
+
"step": 300
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 1.57,
|
| 217 |
+
"learning_rate": 2.634892086330935e-07,
|
| 218 |
+
"logps/chosen": -141.9879913330078,
|
| 219 |
+
"logps/rejected": -107.29058837890625,
|
| 220 |
+
"loss": 0.349,
|
| 221 |
+
"losses/dpo": 0.3294835686683655,
|
| 222 |
+
"losses/sft": 1.2578469514846802,
|
| 223 |
+
"losses/total": 0.3294835686683655,
|
| 224 |
+
"ref_logps/chosen": -152.93646240234375,
|
| 225 |
+
"ref_logps/rejected": -105.73262786865234,
|
| 226 |
+
"rewards/accuracies": 0.8819999694824219,
|
| 227 |
+
"rewards/chosen": 1.0948452949523926,
|
| 228 |
+
"rewards/margins": 1.2506405115127563,
|
| 229 |
+
"rewards/rejected": -0.15579518675804138,
|
| 230 |
+
"step": 325
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.69,
|
| 234 |
+
"learning_rate": 2.410071942446043e-07,
|
| 235 |
+
"logps/chosen": -142.55641174316406,
|
| 236 |
+
"logps/rejected": -108.43463134765625,
|
| 237 |
+
"loss": 0.3091,
|
| 238 |
+
"losses/dpo": 0.3135191798210144,
|
| 239 |
+
"losses/sft": 1.2662179470062256,
|
| 240 |
+
"losses/total": 0.3135191798210144,
|
| 241 |
+
"ref_logps/chosen": -154.17462158203125,
|
| 242 |
+
"ref_logps/rejected": -105.654052734375,
|
| 243 |
+
"rewards/accuracies": 0.8995000720024109,
|
| 244 |
+
"rewards/chosen": 1.1618221998214722,
|
| 245 |
+
"rewards/margins": 1.4398791790008545,
|
| 246 |
+
"rewards/rejected": -0.2780568599700928,
|
| 247 |
+
"step": 350
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 1.82,
|
| 251 |
+
"learning_rate": 2.185251798561151e-07,
|
| 252 |
+
"logps/chosen": -141.23895263671875,
|
| 253 |
+
"logps/rejected": -108.13822937011719,
|
| 254 |
+
"loss": 0.3071,
|
| 255 |
+
"losses/dpo": 0.299526184797287,
|
| 256 |
+
"losses/sft": 1.2248637676239014,
|
| 257 |
+
"losses/total": 0.299526184797287,
|
| 258 |
+
"ref_logps/chosen": -153.57838439941406,
|
| 259 |
+
"ref_logps/rejected": -105.29540252685547,
|
| 260 |
+
"rewards/accuracies": 0.8889999389648438,
|
| 261 |
+
"rewards/chosen": 1.2339427471160889,
|
| 262 |
+
"rewards/margins": 1.5182260274887085,
|
| 263 |
+
"rewards/rejected": -0.28428351879119873,
|
| 264 |
+
"step": 375
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.94,
|
| 268 |
+
"learning_rate": 1.960431654676259e-07,
|
| 269 |
+
"logps/chosen": -143.84800720214844,
|
| 270 |
+
"logps/rejected": -109.23783111572266,
|
| 271 |
+
"loss": 0.3079,
|
| 272 |
+
"losses/dpo": 0.31919562816619873,
|
| 273 |
+
"losses/sft": 1.2433806657791138,
|
| 274 |
+
"losses/total": 0.31919562816619873,
|
| 275 |
+
"ref_logps/chosen": -155.57464599609375,
|
| 276 |
+
"ref_logps/rejected": -105.85549926757812,
|
| 277 |
+
"rewards/accuracies": 0.8924999237060547,
|
| 278 |
+
"rewards/chosen": 1.1726653575897217,
|
| 279 |
+
"rewards/margins": 1.510898470878601,
|
| 280 |
+
"rewards/rejected": -0.33823302388191223,
|
| 281 |
+
"step": 400
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.06,
|
| 285 |
+
"learning_rate": 1.735611510791367e-07,
|
| 286 |
+
"logps/chosen": -142.04263305664062,
|
| 287 |
+
"logps/rejected": -110.5551528930664,
|
| 288 |
+
"loss": 0.2964,
|
| 289 |
+
"losses/dpo": 0.29985514283180237,
|
| 290 |
+
"losses/sft": 1.256807565689087,
|
| 291 |
+
"losses/total": 0.29985514283180237,
|
| 292 |
+
"ref_logps/chosen": -153.96051025390625,
|
| 293 |
+
"ref_logps/rejected": -106.5252914428711,
|
| 294 |
+
"rewards/accuracies": 0.8864999413490295,
|
| 295 |
+
"rewards/chosen": 1.1917892694473267,
|
| 296 |
+
"rewards/margins": 1.5947766304016113,
|
| 297 |
+
"rewards/rejected": -0.4029873013496399,
|
| 298 |
+
"step": 425
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 2.18,
|
| 302 |
+
"learning_rate": 1.5107913669064747e-07,
|
| 303 |
+
"logps/chosen": -141.11489868164062,
|
| 304 |
+
"logps/rejected": -110.66223907470703,
|
| 305 |
+
"loss": 0.2921,
|
| 306 |
+
"losses/dpo": 0.29900702834129333,
|
| 307 |
+
"losses/sft": 1.223684549331665,
|
| 308 |
+
"losses/total": 0.29900702834129333,
|
| 309 |
+
"ref_logps/chosen": -153.32489013671875,
|
| 310 |
+
"ref_logps/rejected": -106.36825561523438,
|
| 311 |
+
"rewards/accuracies": 0.9029999375343323,
|
| 312 |
+
"rewards/chosen": 1.2209986448287964,
|
| 313 |
+
"rewards/margins": 1.6503956317901611,
|
| 314 |
+
"rewards/rejected": -0.42939692735671997,
|
| 315 |
+
"step": 450
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 2.3,
|
| 319 |
+
"learning_rate": 1.2859712230215827e-07,
|
| 320 |
+
"logps/chosen": -140.06715393066406,
|
| 321 |
+
"logps/rejected": -111.3785629272461,
|
| 322 |
+
"loss": 0.2771,
|
| 323 |
+
"losses/dpo": 0.2982315719127655,
|
| 324 |
+
"losses/sft": 1.2262225151062012,
|
| 325 |
+
"losses/total": 0.2982315719127655,
|
| 326 |
+
"ref_logps/chosen": -152.20010375976562,
|
| 327 |
+
"ref_logps/rejected": -106.2849349975586,
|
| 328 |
+
"rewards/accuracies": 0.8995000720024109,
|
| 329 |
+
"rewards/chosen": 1.2132951021194458,
|
| 330 |
+
"rewards/margins": 1.722659707069397,
|
| 331 |
+
"rewards/rejected": -0.509364664554596,
|
| 332 |
+
"step": 475
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.42,
|
| 336 |
+
"learning_rate": 1.0611510791366907e-07,
|
| 337 |
+
"logps/chosen": -142.41409301757812,
|
| 338 |
+
"logps/rejected": -112.02964782714844,
|
| 339 |
+
"loss": 0.2692,
|
| 340 |
+
"losses/dpo": 0.26506972312927246,
|
| 341 |
+
"losses/sft": 1.2326300144195557,
|
| 342 |
+
"losses/total": 0.26506972312927246,
|
| 343 |
+
"ref_logps/chosen": -154.93585205078125,
|
| 344 |
+
"ref_logps/rejected": -106.74388885498047,
|
| 345 |
+
"rewards/accuracies": 0.9050000905990601,
|
| 346 |
+
"rewards/chosen": 1.252176284790039,
|
| 347 |
+
"rewards/margins": 1.7807520627975464,
|
| 348 |
+
"rewards/rejected": -0.5285759568214417,
|
| 349 |
+
"step": 500
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 2.54,
|
| 353 |
+
"learning_rate": 8.363309352517984e-08,
|
| 354 |
+
"logps/chosen": -141.35763549804688,
|
| 355 |
+
"logps/rejected": -111.21255493164062,
|
| 356 |
+
"loss": 0.2639,
|
| 357 |
+
"losses/dpo": 0.2448505461215973,
|
| 358 |
+
"losses/sft": 1.2256782054901123,
|
| 359 |
+
"losses/total": 0.2448505461215973,
|
| 360 |
+
"ref_logps/chosen": -153.86410522460938,
|
| 361 |
+
"ref_logps/rejected": -105.44302368164062,
|
| 362 |
+
"rewards/accuracies": 0.9070001244544983,
|
| 363 |
+
"rewards/chosen": 1.250647783279419,
|
| 364 |
+
"rewards/margins": 1.8276008367538452,
|
| 365 |
+
"rewards/rejected": -0.576953113079071,
|
| 366 |
+
"step": 525
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 2.66,
|
| 370 |
+
"learning_rate": 6.115107913669064e-08,
|
| 371 |
+
"logps/chosen": -140.52743530273438,
|
| 372 |
+
"logps/rejected": -111.19217681884766,
|
| 373 |
+
"loss": 0.253,
|
| 374 |
+
"losses/dpo": 0.2362019419670105,
|
| 375 |
+
"losses/sft": 1.219558835029602,
|
| 376 |
+
"losses/total": 0.2362019419670105,
|
| 377 |
+
"ref_logps/chosen": -153.1510009765625,
|
| 378 |
+
"ref_logps/rejected": -104.75523376464844,
|
| 379 |
+
"rewards/accuracies": 0.925000011920929,
|
| 380 |
+
"rewards/chosen": 1.2623597383499146,
|
| 381 |
+
"rewards/margins": 1.9060544967651367,
|
| 382 |
+
"rewards/rejected": -0.6436949372291565,
|
| 383 |
+
"step": 550
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 2.78,
|
| 387 |
+
"learning_rate": 3.866906474820144e-08,
|
| 388 |
+
"logps/chosen": -142.2875213623047,
|
| 389 |
+
"logps/rejected": -113.81513977050781,
|
| 390 |
+
"loss": 0.2529,
|
| 391 |
+
"losses/dpo": 0.24827653169631958,
|
| 392 |
+
"losses/sft": 1.2383555173873901,
|
| 393 |
+
"losses/total": 0.24827653169631958,
|
| 394 |
+
"ref_logps/chosen": -154.96365356445312,
|
| 395 |
+
"ref_logps/rejected": -107.25000762939453,
|
| 396 |
+
"rewards/accuracies": 0.9150000214576721,
|
| 397 |
+
"rewards/chosen": 1.2676140069961548,
|
| 398 |
+
"rewards/margins": 1.9241262674331665,
|
| 399 |
+
"rewards/rejected": -0.656512439250946,
|
| 400 |
+
"step": 575
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 2.91,
|
| 404 |
+
"learning_rate": 1.618705035971223e-08,
|
| 405 |
+
"logps/chosen": -140.9930877685547,
|
| 406 |
+
"logps/rejected": -113.36444091796875,
|
| 407 |
+
"loss": 0.2651,
|
| 408 |
+
"losses/dpo": 0.27665671706199646,
|
| 409 |
+
"losses/sft": 1.2188746929168701,
|
| 410 |
+
"losses/total": 0.27665671706199646,
|
| 411 |
+
"ref_logps/chosen": -153.5058135986328,
|
| 412 |
+
"ref_logps/rejected": -107.16682434082031,
|
| 413 |
+
"rewards/accuracies": 0.9030001163482666,
|
| 414 |
+
"rewards/chosen": 1.251274585723877,
|
| 415 |
+
"rewards/margins": 1.871036410331726,
|
| 416 |
+
"rewards/rejected": -0.6197616457939148,
|
| 417 |
+
"step": 600
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 2.99,
|
| 421 |
+
"step": 618,
|
| 422 |
+
"total_flos": 0.0,
|
| 423 |
+
"train_loss": 0.41181609237078326,
|
| 424 |
+
"train_runtime": 7630.1245,
|
| 425 |
+
"train_samples_per_second": 6.494,
|
| 426 |
+
"train_steps_per_second": 0.081
|
| 427 |
+
}
|
| 428 |
+
],
|
| 429 |
+
"logging_steps": 25,
|
| 430 |
+
"max_steps": 618,
|
| 431 |
+
"num_input_tokens_seen": 0,
|
| 432 |
+
"num_train_epochs": 3,
|
| 433 |
+
"save_steps": 500,
|
| 434 |
+
"stateful_callbacks": {},
|
| 435 |
+
"total_flos": 0.0,
|
| 436 |
+
"train_batch_size": 4,
|
| 437 |
+
"trial_name": null,
|
| 438 |
+
"trial_params": null
|
| 439 |
+
}
|