Upload folder using huggingface_hub
Browse files- 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 +847 -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 0.4.0
|
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 |
+
"k_proj",
|
| 18 |
+
"v_proj",
|
| 19 |
+
"o_proj",
|
| 20 |
+
"q_proj",
|
| 21 |
+
"out_proj",
|
| 22 |
+
"fc1",
|
| 23 |
+
"down_proj",
|
| 24 |
+
"fc2",
|
| 25 |
+
"gate_proj",
|
| 26 |
+
"up_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 sha256:6ed82dac1b34aae060b966febf7813ff0ae7175570bf89edc6b268991ab480d2
|
| 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 +1,847 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1200,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.06,
|
| 13 |
+
"learning_rate": 1.0416666666666667e-07,
|
| 14 |
+
"logps/chosen": -50.84611511230469,
|
| 15 |
+
"logps/rejected": -68.16911315917969,
|
| 16 |
+
"loss": 0.7128,
|
| 17 |
+
"losses/dpo": 0.7152940630912781,
|
| 18 |
+
"losses/sft": 1.772851586341858,
|
| 19 |
+
"losses/total": 0.7152940630912781,
|
| 20 |
+
"ref_logps/chosen": -50.76835632324219,
|
| 21 |
+
"ref_logps/rejected": -68.15713500976562,
|
| 22 |
+
"rewards/accuracies": 0.4920000433921814,
|
| 23 |
+
"rewards/chosen": -0.007775493897497654,
|
| 24 |
+
"rewards/margins": -0.006577868480235338,
|
| 25 |
+
"rewards/rejected": -0.0011976259993389249,
|
| 26 |
+
"step": 25
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.12,
|
| 30 |
+
"learning_rate": 2.0833333333333333e-07,
|
| 31 |
+
"logps/chosen": -51.992218017578125,
|
| 32 |
+
"logps/rejected": -67.8126220703125,
|
| 33 |
+
"loss": 0.7059,
|
| 34 |
+
"losses/dpo": 0.7131974697113037,
|
| 35 |
+
"losses/sft": 1.781252384185791,
|
| 36 |
+
"losses/total": 0.7131974697113037,
|
| 37 |
+
"ref_logps/chosen": -52.02339172363281,
|
| 38 |
+
"ref_logps/rejected": -67.7878646850586,
|
| 39 |
+
"rewards/accuracies": 0.5110000371932983,
|
| 40 |
+
"rewards/chosen": 0.003117597894743085,
|
| 41 |
+
"rewards/margins": 0.005593638867139816,
|
| 42 |
+
"rewards/rejected": -0.0024760416708886623,
|
| 43 |
+
"step": 50
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.19,
|
| 47 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 48 |
+
"logps/chosen": -52.34601593017578,
|
| 49 |
+
"logps/rejected": -67.67086029052734,
|
| 50 |
+
"loss": 0.7097,
|
| 51 |
+
"losses/dpo": 0.7076306939125061,
|
| 52 |
+
"losses/sft": 1.7903097867965698,
|
| 53 |
+
"losses/total": 0.7076306939125061,
|
| 54 |
+
"ref_logps/chosen": -52.40769577026367,
|
| 55 |
+
"ref_logps/rejected": -67.70091247558594,
|
| 56 |
+
"rewards/accuracies": 0.49650001525878906,
|
| 57 |
+
"rewards/chosen": 0.006168678868561983,
|
| 58 |
+
"rewards/margins": 0.003163211978971958,
|
| 59 |
+
"rewards/rejected": 0.003005466191098094,
|
| 60 |
+
"step": 75
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.25,
|
| 64 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 65 |
+
"logps/chosen": -51.974483489990234,
|
| 66 |
+
"logps/rejected": -68.44440460205078,
|
| 67 |
+
"loss": 0.7115,
|
| 68 |
+
"losses/dpo": 0.7194903492927551,
|
| 69 |
+
"losses/sft": 1.777998924255371,
|
| 70 |
+
"losses/total": 0.7194903492927551,
|
| 71 |
+
"ref_logps/chosen": -51.95515441894531,
|
| 72 |
+
"ref_logps/rejected": -68.4430923461914,
|
| 73 |
+
"rewards/accuracies": 0.4794999957084656,
|
| 74 |
+
"rewards/chosen": -0.0019327354384586215,
|
| 75 |
+
"rewards/margins": -0.0018010810017585754,
|
| 76 |
+
"rewards/rejected": -0.00013165354903321713,
|
| 77 |
+
"step": 100
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 0.31,
|
| 81 |
+
"learning_rate": 4.976851851851852e-07,
|
| 82 |
+
"logps/chosen": -51.21366882324219,
|
| 83 |
+
"logps/rejected": -68.10318756103516,
|
| 84 |
+
"loss": 0.7096,
|
| 85 |
+
"losses/dpo": 0.7016645073890686,
|
| 86 |
+
"losses/sft": 1.7784287929534912,
|
| 87 |
+
"losses/total": 0.7016645073890686,
|
| 88 |
+
"ref_logps/chosen": -51.175315856933594,
|
| 89 |
+
"ref_logps/rejected": -68.03705596923828,
|
| 90 |
+
"rewards/accuracies": 0.5074999928474426,
|
| 91 |
+
"rewards/chosen": -0.003834722563624382,
|
| 92 |
+
"rewards/margins": 0.002778256544843316,
|
| 93 |
+
"rewards/rejected": -0.006612981669604778,
|
| 94 |
+
"step": 125
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.38,
|
| 98 |
+
"learning_rate": 4.861111111111111e-07,
|
| 99 |
+
"logps/chosen": -49.84953689575195,
|
| 100 |
+
"logps/rejected": -66.59363555908203,
|
| 101 |
+
"loss": 0.6989,
|
| 102 |
+
"losses/dpo": 0.6881049275398254,
|
| 103 |
+
"losses/sft": 1.7335822582244873,
|
| 104 |
+
"losses/total": 0.6881049275398254,
|
| 105 |
+
"ref_logps/chosen": -49.83342361450195,
|
| 106 |
+
"ref_logps/rejected": -66.36032104492188,
|
| 107 |
+
"rewards/accuracies": 0.5224999785423279,
|
| 108 |
+
"rewards/chosen": -0.0016115426551550627,
|
| 109 |
+
"rewards/margins": 0.021719880402088165,
|
| 110 |
+
"rewards/rejected": -0.023331422358751297,
|
| 111 |
+
"step": 150
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.44,
|
| 115 |
+
"learning_rate": 4.74537037037037e-07,
|
| 116 |
+
"logps/chosen": -51.20610427856445,
|
| 117 |
+
"logps/rejected": -68.37096405029297,
|
| 118 |
+
"loss": 0.6965,
|
| 119 |
+
"losses/dpo": 0.6989789605140686,
|
| 120 |
+
"losses/sft": 1.8248052597045898,
|
| 121 |
+
"losses/total": 0.6989789605140686,
|
| 122 |
+
"ref_logps/chosen": -51.07723617553711,
|
| 123 |
+
"ref_logps/rejected": -67.93907165527344,
|
| 124 |
+
"rewards/accuracies": 0.5259999632835388,
|
| 125 |
+
"rewards/chosen": -0.012886175885796547,
|
| 126 |
+
"rewards/margins": 0.0303032249212265,
|
| 127 |
+
"rewards/rejected": -0.0431894026696682,
|
| 128 |
+
"step": 175
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.5,
|
| 132 |
+
"learning_rate": 4.6296296296296297e-07,
|
| 133 |
+
"logps/chosen": -53.1021842956543,
|
| 134 |
+
"logps/rejected": -69.64921569824219,
|
| 135 |
+
"loss": 0.6821,
|
| 136 |
+
"losses/dpo": 0.6905575394630432,
|
| 137 |
+
"losses/sft": 1.7901018857955933,
|
| 138 |
+
"losses/total": 0.6905575394630432,
|
| 139 |
+
"ref_logps/chosen": -52.89280319213867,
|
| 140 |
+
"ref_logps/rejected": -68.83776092529297,
|
| 141 |
+
"rewards/accuracies": 0.5610000491142273,
|
| 142 |
+
"rewards/chosen": -0.02093815989792347,
|
| 143 |
+
"rewards/margins": 0.0602080263197422,
|
| 144 |
+
"rewards/rejected": -0.08114618062973022,
|
| 145 |
+
"step": 200
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.56,
|
| 149 |
+
"learning_rate": 4.513888888888889e-07,
|
| 150 |
+
"logps/chosen": -51.60838317871094,
|
| 151 |
+
"logps/rejected": -68.19721984863281,
|
| 152 |
+
"loss": 0.6783,
|
| 153 |
+
"losses/dpo": 0.6758668422698975,
|
| 154 |
+
"losses/sft": 1.7487859725952148,
|
| 155 |
+
"losses/total": 0.6758668422698975,
|
| 156 |
+
"ref_logps/chosen": -51.219181060791016,
|
| 157 |
+
"ref_logps/rejected": -67.0694580078125,
|
| 158 |
+
"rewards/accuracies": 0.5529999732971191,
|
| 159 |
+
"rewards/chosen": -0.03891964256763458,
|
| 160 |
+
"rewards/margins": 0.0738559141755104,
|
| 161 |
+
"rewards/rejected": -0.11277556419372559,
|
| 162 |
+
"step": 225
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.62,
|
| 166 |
+
"learning_rate": 4.398148148148148e-07,
|
| 167 |
+
"logps/chosen": -51.5633544921875,
|
| 168 |
+
"logps/rejected": -70.04096221923828,
|
| 169 |
+
"loss": 0.6606,
|
| 170 |
+
"losses/dpo": 0.6762534976005554,
|
| 171 |
+
"losses/sft": 1.7945468425750732,
|
| 172 |
+
"losses/total": 0.6762534976005554,
|
| 173 |
+
"ref_logps/chosen": -51.164485931396484,
|
| 174 |
+
"ref_logps/rejected": -68.47702026367188,
|
| 175 |
+
"rewards/accuracies": 0.5925000309944153,
|
| 176 |
+
"rewards/chosen": -0.03988657891750336,
|
| 177 |
+
"rewards/margins": 0.11650750786066055,
|
| 178 |
+
"rewards/rejected": -0.1563940793275833,
|
| 179 |
+
"step": 250
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 0.69,
|
| 183 |
+
"learning_rate": 4.2824074074074075e-07,
|
| 184 |
+
"logps/chosen": -52.760066986083984,
|
| 185 |
+
"logps/rejected": -71.27080535888672,
|
| 186 |
+
"loss": 0.6541,
|
| 187 |
+
"losses/dpo": 0.6559135317802429,
|
| 188 |
+
"losses/sft": 1.8313995599746704,
|
| 189 |
+
"losses/total": 0.6559135317802429,
|
| 190 |
+
"ref_logps/chosen": -52.14683151245117,
|
| 191 |
+
"ref_logps/rejected": -69.2181625366211,
|
| 192 |
+
"rewards/accuracies": 0.6005000472068787,
|
| 193 |
+
"rewards/chosen": -0.0613233856856823,
|
| 194 |
+
"rewards/margins": 0.14394104480743408,
|
| 195 |
+
"rewards/rejected": -0.20526443421840668,
|
| 196 |
+
"step": 275
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.75,
|
| 200 |
+
"learning_rate": 4.1666666666666667e-07,
|
| 201 |
+
"logps/chosen": -53.022369384765625,
|
| 202 |
+
"logps/rejected": -70.22279357910156,
|
| 203 |
+
"loss": 0.6444,
|
| 204 |
+
"losses/dpo": 0.6332589387893677,
|
| 205 |
+
"losses/sft": 1.8123092651367188,
|
| 206 |
+
"losses/total": 0.6332589387893677,
|
| 207 |
+
"ref_logps/chosen": -52.272945404052734,
|
| 208 |
+
"ref_logps/rejected": -67.84944152832031,
|
| 209 |
+
"rewards/accuracies": 0.6039999723434448,
|
| 210 |
+
"rewards/chosen": -0.07494210451841354,
|
| 211 |
+
"rewards/margins": 0.16239266097545624,
|
| 212 |
+
"rewards/rejected": -0.2373347282409668,
|
| 213 |
+
"step": 300
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.81,
|
| 217 |
+
"learning_rate": 4.050925925925926e-07,
|
| 218 |
+
"logps/chosen": -52.53852462768555,
|
| 219 |
+
"logps/rejected": -69.21473693847656,
|
| 220 |
+
"loss": 0.6472,
|
| 221 |
+
"losses/dpo": 0.6568192839622498,
|
| 222 |
+
"losses/sft": 1.808381199836731,
|
| 223 |
+
"losses/total": 0.6568192839622498,
|
| 224 |
+
"ref_logps/chosen": -51.50153350830078,
|
| 225 |
+
"ref_logps/rejected": -66.50575256347656,
|
| 226 |
+
"rewards/accuracies": 0.6184999942779541,
|
| 227 |
+
"rewards/chosen": -0.10369860380887985,
|
| 228 |
+
"rewards/margins": 0.1671994924545288,
|
| 229 |
+
"rewards/rejected": -0.27089810371398926,
|
| 230 |
+
"step": 325
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.88,
|
| 234 |
+
"learning_rate": 3.9351851851851854e-07,
|
| 235 |
+
"logps/chosen": -52.68033218383789,
|
| 236 |
+
"logps/rejected": -69.9216079711914,
|
| 237 |
+
"loss": 0.6322,
|
| 238 |
+
"losses/dpo": 0.6414695978164673,
|
| 239 |
+
"losses/sft": 1.7517987489700317,
|
| 240 |
+
"losses/total": 0.6414695978164673,
|
| 241 |
+
"ref_logps/chosen": -51.4389533996582,
|
| 242 |
+
"ref_logps/rejected": -66.67315673828125,
|
| 243 |
+
"rewards/accuracies": 0.6355000138282776,
|
| 244 |
+
"rewards/chosen": -0.12413790076971054,
|
| 245 |
+
"rewards/margins": 0.20070795714855194,
|
| 246 |
+
"rewards/rejected": -0.3248458802700043,
|
| 247 |
+
"step": 350
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.94,
|
| 251 |
+
"learning_rate": 3.819444444444444e-07,
|
| 252 |
+
"logps/chosen": -53.01972579956055,
|
| 253 |
+
"logps/rejected": -69.83106231689453,
|
| 254 |
+
"loss": 0.6254,
|
| 255 |
+
"losses/dpo": 0.6416024565696716,
|
| 256 |
+
"losses/sft": 1.806755781173706,
|
| 257 |
+
"losses/total": 0.6416024565696716,
|
| 258 |
+
"ref_logps/chosen": -51.760948181152344,
|
| 259 |
+
"ref_logps/rejected": -66.29093170166016,
|
| 260 |
+
"rewards/accuracies": 0.6260000467300415,
|
| 261 |
+
"rewards/chosen": -0.12587812542915344,
|
| 262 |
+
"rewards/margins": 0.22813528776168823,
|
| 263 |
+
"rewards/rejected": -0.3540134131908417,
|
| 264 |
+
"step": 375
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 1.0,
|
| 268 |
+
"learning_rate": 3.703703703703703e-07,
|
| 269 |
+
"logps/chosen": -53.866825103759766,
|
| 270 |
+
"logps/rejected": -73.0006332397461,
|
| 271 |
+
"loss": 0.6223,
|
| 272 |
+
"losses/dpo": 0.6281777024269104,
|
| 273 |
+
"losses/sft": 1.8359609842300415,
|
| 274 |
+
"losses/total": 0.6281777024269104,
|
| 275 |
+
"ref_logps/chosen": -52.29502868652344,
|
| 276 |
+
"ref_logps/rejected": -68.76205444335938,
|
| 277 |
+
"rewards/accuracies": 0.624500036239624,
|
| 278 |
+
"rewards/chosen": -0.15717901289463043,
|
| 279 |
+
"rewards/margins": 0.26667752861976624,
|
| 280 |
+
"rewards/rejected": -0.42385655641555786,
|
| 281 |
+
"step": 400
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.06,
|
| 285 |
+
"learning_rate": 3.587962962962963e-07,
|
| 286 |
+
"logps/chosen": -52.83345031738281,
|
| 287 |
+
"logps/rejected": -73.19969940185547,
|
| 288 |
+
"loss": 0.6044,
|
| 289 |
+
"losses/dpo": 0.6066713929176331,
|
| 290 |
+
"losses/sft": 1.8482356071472168,
|
| 291 |
+
"losses/total": 0.6066713929176331,
|
| 292 |
+
"ref_logps/chosen": -51.23924255371094,
|
| 293 |
+
"ref_logps/rejected": -68.46554565429688,
|
| 294 |
+
"rewards/accuracies": 0.658500075340271,
|
| 295 |
+
"rewards/chosen": -0.15942102670669556,
|
| 296 |
+
"rewards/margins": 0.31399470567703247,
|
| 297 |
+
"rewards/rejected": -0.4734157621860504,
|
| 298 |
+
"step": 425
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.12,
|
| 302 |
+
"learning_rate": 3.472222222222222e-07,
|
| 303 |
+
"logps/chosen": -53.00102996826172,
|
| 304 |
+
"logps/rejected": -71.68830871582031,
|
| 305 |
+
"loss": 0.6107,
|
| 306 |
+
"losses/dpo": 0.5940313935279846,
|
| 307 |
+
"losses/sft": 1.8907349109649658,
|
| 308 |
+
"losses/total": 0.5940313935279846,
|
| 309 |
+
"ref_logps/chosen": -51.17875671386719,
|
| 310 |
+
"ref_logps/rejected": -66.69224548339844,
|
| 311 |
+
"rewards/accuracies": 0.6355000138282776,
|
| 312 |
+
"rewards/chosen": -0.18222743272781372,
|
| 313 |
+
"rewards/margins": 0.3173789978027344,
|
| 314 |
+
"rewards/rejected": -0.4996064305305481,
|
| 315 |
+
"step": 450
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.19,
|
| 319 |
+
"learning_rate": 3.3564814814814815e-07,
|
| 320 |
+
"logps/chosen": -54.13747024536133,
|
| 321 |
+
"logps/rejected": -75.39285278320312,
|
| 322 |
+
"loss": 0.5906,
|
| 323 |
+
"losses/dpo": 0.6001964807510376,
|
| 324 |
+
"losses/sft": 1.837428331375122,
|
| 325 |
+
"losses/total": 0.6001964807510376,
|
| 326 |
+
"ref_logps/chosen": -52.013328552246094,
|
| 327 |
+
"ref_logps/rejected": -69.39274597167969,
|
| 328 |
+
"rewards/accuracies": 0.6559999585151672,
|
| 329 |
+
"rewards/chosen": -0.21241408586502075,
|
| 330 |
+
"rewards/margins": 0.387596994638443,
|
| 331 |
+
"rewards/rejected": -0.6000111103057861,
|
| 332 |
+
"step": 475
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.25,
|
| 336 |
+
"learning_rate": 3.2407407407407406e-07,
|
| 337 |
+
"logps/chosen": -52.84211730957031,
|
| 338 |
+
"logps/rejected": -73.44519805908203,
|
| 339 |
+
"loss": 0.588,
|
| 340 |
+
"losses/dpo": 0.5720361471176147,
|
| 341 |
+
"losses/sft": 1.82402765750885,
|
| 342 |
+
"losses/total": 0.5720361471176147,
|
| 343 |
+
"ref_logps/chosen": -50.860145568847656,
|
| 344 |
+
"ref_logps/rejected": -67.45731353759766,
|
| 345 |
+
"rewards/accuracies": 0.6580000519752502,
|
| 346 |
+
"rewards/chosen": -0.1981971114873886,
|
| 347 |
+
"rewards/margins": 0.40059128403663635,
|
| 348 |
+
"rewards/rejected": -0.5987883806228638,
|
| 349 |
+
"step": 500
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 1.31,
|
| 353 |
+
"learning_rate": 3.1249999999999997e-07,
|
| 354 |
+
"logps/chosen": -54.73566436767578,
|
| 355 |
+
"logps/rejected": -73.35672760009766,
|
| 356 |
+
"loss": 0.5864,
|
| 357 |
+
"losses/dpo": 0.5692346096038818,
|
| 358 |
+
"losses/sft": 1.8467037677764893,
|
| 359 |
+
"losses/total": 0.5692346096038818,
|
| 360 |
+
"ref_logps/chosen": -52.32421875,
|
| 361 |
+
"ref_logps/rejected": -66.84322357177734,
|
| 362 |
+
"rewards/accuracies": 0.6690000295639038,
|
| 363 |
+
"rewards/chosen": -0.24114486575126648,
|
| 364 |
+
"rewards/margins": 0.4102060794830322,
|
| 365 |
+
"rewards/rejected": -0.6513509154319763,
|
| 366 |
+
"step": 525
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 1.38,
|
| 370 |
+
"learning_rate": 3.0092592592592594e-07,
|
| 371 |
+
"logps/chosen": -54.00200653076172,
|
| 372 |
+
"logps/rejected": -77.5279312133789,
|
| 373 |
+
"loss": 0.5712,
|
| 374 |
+
"losses/dpo": 0.5749830007553101,
|
| 375 |
+
"losses/sft": 1.8555569648742676,
|
| 376 |
+
"losses/total": 0.5749830007553101,
|
| 377 |
+
"ref_logps/chosen": -51.46855163574219,
|
| 378 |
+
"ref_logps/rejected": -70.09744262695312,
|
| 379 |
+
"rewards/accuracies": 0.6720000505447388,
|
| 380 |
+
"rewards/chosen": -0.25334587693214417,
|
| 381 |
+
"rewards/margins": 0.48970410227775574,
|
| 382 |
+
"rewards/rejected": -0.7430499196052551,
|
| 383 |
+
"step": 550
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 1.44,
|
| 387 |
+
"learning_rate": 2.8935185185185185e-07,
|
| 388 |
+
"logps/chosen": -53.216957092285156,
|
| 389 |
+
"logps/rejected": -74.7680892944336,
|
| 390 |
+
"loss": 0.5712,
|
| 391 |
+
"losses/dpo": 0.5629280209541321,
|
| 392 |
+
"losses/sft": 1.8214166164398193,
|
| 393 |
+
"losses/total": 0.5629280209541321,
|
| 394 |
+
"ref_logps/chosen": -50.563720703125,
|
| 395 |
+
"ref_logps/rejected": -67.18008422851562,
|
| 396 |
+
"rewards/accuracies": 0.6755000352859497,
|
| 397 |
+
"rewards/chosen": -0.2653237581253052,
|
| 398 |
+
"rewards/margins": 0.4934755265712738,
|
| 399 |
+
"rewards/rejected": -0.7587993144989014,
|
| 400 |
+
"step": 575
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.5,
|
| 404 |
+
"learning_rate": 2.7777777777777776e-07,
|
| 405 |
+
"logps/chosen": -55.059600830078125,
|
| 406 |
+
"logps/rejected": -76.23825073242188,
|
| 407 |
+
"loss": 0.5577,
|
| 408 |
+
"losses/dpo": 0.5477968454360962,
|
| 409 |
+
"losses/sft": 1.9264967441558838,
|
| 410 |
+
"losses/total": 0.5477968454360962,
|
| 411 |
+
"ref_logps/chosen": -52.09434509277344,
|
| 412 |
+
"ref_logps/rejected": -67.781005859375,
|
| 413 |
+
"rewards/accuracies": 0.6880000233650208,
|
| 414 |
+
"rewards/chosen": -0.2965255081653595,
|
| 415 |
+
"rewards/margins": 0.5491979122161865,
|
| 416 |
+
"rewards/rejected": -0.8457233905792236,
|
| 417 |
+
"step": 600
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.56,
|
| 421 |
+
"learning_rate": 2.662037037037037e-07,
|
| 422 |
+
"logps/chosen": -56.8978271484375,
|
| 423 |
+
"logps/rejected": -78.7895736694336,
|
| 424 |
+
"loss": 0.5667,
|
| 425 |
+
"losses/dpo": 0.5594893097877502,
|
| 426 |
+
"losses/sft": 1.9116603136062622,
|
| 427 |
+
"losses/total": 0.5594893097877502,
|
| 428 |
+
"ref_logps/chosen": -53.51485824584961,
|
| 429 |
+
"ref_logps/rejected": -69.69821166992188,
|
| 430 |
+
"rewards/accuracies": 0.6595000624656677,
|
| 431 |
+
"rewards/chosen": -0.338296502828598,
|
| 432 |
+
"rewards/margins": 0.5708397030830383,
|
| 433 |
+
"rewards/rejected": -0.9091362953186035,
|
| 434 |
+
"step": 625
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.62,
|
| 438 |
+
"learning_rate": 2.5462962962962963e-07,
|
| 439 |
+
"logps/chosen": -55.704219818115234,
|
| 440 |
+
"logps/rejected": -77.06362915039062,
|
| 441 |
+
"loss": 0.5734,
|
| 442 |
+
"losses/dpo": 0.583986222743988,
|
| 443 |
+
"losses/sft": 1.8731328248977661,
|
| 444 |
+
"losses/total": 0.583986222743988,
|
| 445 |
+
"ref_logps/chosen": -52.17438888549805,
|
| 446 |
+
"ref_logps/rejected": -67.96855926513672,
|
| 447 |
+
"rewards/accuracies": 0.6725000143051147,
|
| 448 |
+
"rewards/chosen": -0.3529825508594513,
|
| 449 |
+
"rewards/margins": 0.5565236806869507,
|
| 450 |
+
"rewards/rejected": -0.9095062017440796,
|
| 451 |
+
"step": 650
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.69,
|
| 455 |
+
"learning_rate": 2.4305555555555555e-07,
|
| 456 |
+
"logps/chosen": -54.6390266418457,
|
| 457 |
+
"logps/rejected": -76.69291687011719,
|
| 458 |
+
"loss": 0.5568,
|
| 459 |
+
"losses/dpo": 0.5715891718864441,
|
| 460 |
+
"losses/sft": 1.8541003465652466,
|
| 461 |
+
"losses/total": 0.5715891718864441,
|
| 462 |
+
"ref_logps/chosen": -51.134674072265625,
|
| 463 |
+
"ref_logps/rejected": -67.13263702392578,
|
| 464 |
+
"rewards/accuracies": 0.687000036239624,
|
| 465 |
+
"rewards/chosen": -0.3504358232021332,
|
| 466 |
+
"rewards/margins": 0.605591893196106,
|
| 467 |
+
"rewards/rejected": -0.9560276865959167,
|
| 468 |
+
"step": 675
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 1.75,
|
| 472 |
+
"learning_rate": 2.3148148148148148e-07,
|
| 473 |
+
"logps/chosen": -54.50948715209961,
|
| 474 |
+
"logps/rejected": -77.46994018554688,
|
| 475 |
+
"loss": 0.5657,
|
| 476 |
+
"losses/dpo": 0.5799014568328857,
|
| 477 |
+
"losses/sft": 1.8863662481307983,
|
| 478 |
+
"losses/total": 0.5799014568328857,
|
| 479 |
+
"ref_logps/chosen": -50.88352966308594,
|
| 480 |
+
"ref_logps/rejected": -67.8268051147461,
|
| 481 |
+
"rewards/accuracies": 0.6755000352859497,
|
| 482 |
+
"rewards/chosen": -0.36259526014328003,
|
| 483 |
+
"rewards/margins": 0.6017180681228638,
|
| 484 |
+
"rewards/rejected": -0.9643133282661438,
|
| 485 |
+
"step": 700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.81,
|
| 489 |
+
"learning_rate": 2.199074074074074e-07,
|
| 490 |
+
"logps/chosen": -55.129295349121094,
|
| 491 |
+
"logps/rejected": -77.04840087890625,
|
| 492 |
+
"loss": 0.5636,
|
| 493 |
+
"losses/dpo": 0.5747159123420715,
|
| 494 |
+
"losses/sft": 1.926343560218811,
|
| 495 |
+
"losses/total": 0.5747159123420715,
|
| 496 |
+
"ref_logps/chosen": -51.24678039550781,
|
| 497 |
+
"ref_logps/rejected": -67.05614471435547,
|
| 498 |
+
"rewards/accuracies": 0.6730000376701355,
|
| 499 |
+
"rewards/chosen": -0.3882514238357544,
|
| 500 |
+
"rewards/margins": 0.6109741926193237,
|
| 501 |
+
"rewards/rejected": -0.9992256760597229,
|
| 502 |
+
"step": 725
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 1.88,
|
| 506 |
+
"learning_rate": 2.0833333333333333e-07,
|
| 507 |
+
"logps/chosen": -55.22993087768555,
|
| 508 |
+
"logps/rejected": -78.04380798339844,
|
| 509 |
+
"loss": 0.5575,
|
| 510 |
+
"losses/dpo": 0.5746473073959351,
|
| 511 |
+
"losses/sft": 1.8473073244094849,
|
| 512 |
+
"losses/total": 0.5746473073959351,
|
| 513 |
+
"ref_logps/chosen": -51.49522399902344,
|
| 514 |
+
"ref_logps/rejected": -67.86705017089844,
|
| 515 |
+
"rewards/accuracies": 0.6645000576972961,
|
| 516 |
+
"rewards/chosen": -0.3734706938266754,
|
| 517 |
+
"rewards/margins": 0.6442059278488159,
|
| 518 |
+
"rewards/rejected": -1.0176767110824585,
|
| 519 |
+
"step": 750
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 1.94,
|
| 523 |
+
"learning_rate": 1.9675925925925927e-07,
|
| 524 |
+
"logps/chosen": -54.95744323730469,
|
| 525 |
+
"logps/rejected": -77.55332946777344,
|
| 526 |
+
"loss": 0.5428,
|
| 527 |
+
"losses/dpo": 0.5475051999092102,
|
| 528 |
+
"losses/sft": 1.8516862392425537,
|
| 529 |
+
"losses/total": 0.5475051999092102,
|
| 530 |
+
"ref_logps/chosen": -51.13042449951172,
|
| 531 |
+
"ref_logps/rejected": -66.90882110595703,
|
| 532 |
+
"rewards/accuracies": 0.6959999799728394,
|
| 533 |
+
"rewards/chosen": -0.382702112197876,
|
| 534 |
+
"rewards/margins": 0.6817485690116882,
|
| 535 |
+
"rewards/rejected": -1.064450740814209,
|
| 536 |
+
"step": 775
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.0,
|
| 540 |
+
"learning_rate": 1.8518518518518516e-07,
|
| 541 |
+
"logps/chosen": -56.311126708984375,
|
| 542 |
+
"logps/rejected": -79.54733276367188,
|
| 543 |
+
"loss": 0.5453,
|
| 544 |
+
"losses/dpo": 0.529507577419281,
|
| 545 |
+
"losses/sft": 1.9264029264450073,
|
| 546 |
+
"losses/total": 0.529507577419281,
|
| 547 |
+
"ref_logps/chosen": -51.99349594116211,
|
| 548 |
+
"ref_logps/rejected": -68.2928237915039,
|
| 549 |
+
"rewards/accuracies": 0.7039999961853027,
|
| 550 |
+
"rewards/chosen": -0.4317638874053955,
|
| 551 |
+
"rewards/margins": 0.6936854720115662,
|
| 552 |
+
"rewards/rejected": -1.1254491806030273,
|
| 553 |
+
"step": 800
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 2.06,
|
| 557 |
+
"learning_rate": 1.736111111111111e-07,
|
| 558 |
+
"logps/chosen": -55.88136672973633,
|
| 559 |
+
"logps/rejected": -78.89193725585938,
|
| 560 |
+
"loss": 0.5536,
|
| 561 |
+
"losses/dpo": 0.5027932524681091,
|
| 562 |
+
"losses/sft": 1.888191819190979,
|
| 563 |
+
"losses/total": 0.5027932524681091,
|
| 564 |
+
"ref_logps/chosen": -51.50459289550781,
|
| 565 |
+
"ref_logps/rejected": -67.58809661865234,
|
| 566 |
+
"rewards/accuracies": 0.7039999961853027,
|
| 567 |
+
"rewards/chosen": -0.43767789006233215,
|
| 568 |
+
"rewards/margins": 0.6927057504653931,
|
| 569 |
+
"rewards/rejected": -1.1303836107254028,
|
| 570 |
+
"step": 825
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 2.12,
|
| 574 |
+
"learning_rate": 1.6203703703703703e-07,
|
| 575 |
+
"logps/chosen": -56.851905822753906,
|
| 576 |
+
"logps/rejected": -79.84717559814453,
|
| 577 |
+
"loss": 0.5519,
|
| 578 |
+
"losses/dpo": 0.5627533197402954,
|
| 579 |
+
"losses/sft": 1.950637936592102,
|
| 580 |
+
"losses/total": 0.5627533197402954,
|
| 581 |
+
"ref_logps/chosen": -52.47712326049805,
|
| 582 |
+
"ref_logps/rejected": -68.6231918334961,
|
| 583 |
+
"rewards/accuracies": 0.6880000233650208,
|
| 584 |
+
"rewards/chosen": -0.4374789893627167,
|
| 585 |
+
"rewards/margins": 0.6849193572998047,
|
| 586 |
+
"rewards/rejected": -1.1223982572555542,
|
| 587 |
+
"step": 850
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.19,
|
| 591 |
+
"learning_rate": 1.5046296296296297e-07,
|
| 592 |
+
"logps/chosen": -56.792240142822266,
|
| 593 |
+
"logps/rejected": -80.210205078125,
|
| 594 |
+
"loss": 0.5433,
|
| 595 |
+
"losses/dpo": 0.562707245349884,
|
| 596 |
+
"losses/sft": 1.9036242961883545,
|
| 597 |
+
"losses/total": 0.562707245349884,
|
| 598 |
+
"ref_logps/chosen": -52.128883361816406,
|
| 599 |
+
"ref_logps/rejected": -68.35295104980469,
|
| 600 |
+
"rewards/accuracies": 0.6974999904632568,
|
| 601 |
+
"rewards/chosen": -0.46633559465408325,
|
| 602 |
+
"rewards/margins": 0.719389796257019,
|
| 603 |
+
"rewards/rejected": -1.185725450515747,
|
| 604 |
+
"step": 875
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 2.25,
|
| 608 |
+
"learning_rate": 1.3888888888888888e-07,
|
| 609 |
+
"logps/chosen": -56.366416931152344,
|
| 610 |
+
"logps/rejected": -81.07738494873047,
|
| 611 |
+
"loss": 0.5341,
|
| 612 |
+
"losses/dpo": 0.5471935868263245,
|
| 613 |
+
"losses/sft": 1.9060190916061401,
|
| 614 |
+
"losses/total": 0.5471935868263245,
|
| 615 |
+
"ref_logps/chosen": -51.77671432495117,
|
| 616 |
+
"ref_logps/rejected": -68.86495971679688,
|
| 617 |
+
"rewards/accuracies": 0.6914999485015869,
|
| 618 |
+
"rewards/chosen": -0.45896950364112854,
|
| 619 |
+
"rewards/margins": 0.7622722387313843,
|
| 620 |
+
"rewards/rejected": -1.2212417125701904,
|
| 621 |
+
"step": 900
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.31,
|
| 625 |
+
"learning_rate": 1.2731481481481482e-07,
|
| 626 |
+
"logps/chosen": -56.640018463134766,
|
| 627 |
+
"logps/rejected": -81.28205871582031,
|
| 628 |
+
"loss": 0.5195,
|
| 629 |
+
"losses/dpo": 0.5150865316390991,
|
| 630 |
+
"losses/sft": 1.918342113494873,
|
| 631 |
+
"losses/total": 0.5150865316390991,
|
| 632 |
+
"ref_logps/chosen": -52.07315444946289,
|
| 633 |
+
"ref_logps/rejected": -68.57891082763672,
|
| 634 |
+
"rewards/accuracies": 0.7275000214576721,
|
| 635 |
+
"rewards/chosen": -0.4566872715950012,
|
| 636 |
+
"rewards/margins": 0.8136279582977295,
|
| 637 |
+
"rewards/rejected": -1.2703152894973755,
|
| 638 |
+
"step": 925
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.38,
|
| 642 |
+
"learning_rate": 1.1574074074074074e-07,
|
| 643 |
+
"logps/chosen": -56.610965728759766,
|
| 644 |
+
"logps/rejected": -82.6984634399414,
|
| 645 |
+
"loss": 0.5267,
|
| 646 |
+
"losses/dpo": 0.5297408103942871,
|
| 647 |
+
"losses/sft": 1.9613755941390991,
|
| 648 |
+
"losses/total": 0.5297408103942871,
|
| 649 |
+
"ref_logps/chosen": -51.79991149902344,
|
| 650 |
+
"ref_logps/rejected": -69.79439544677734,
|
| 651 |
+
"rewards/accuracies": 0.7074999809265137,
|
| 652 |
+
"rewards/chosen": -0.4811055660247803,
|
| 653 |
+
"rewards/margins": 0.8093020915985107,
|
| 654 |
+
"rewards/rejected": -1.2904075384140015,
|
| 655 |
+
"step": 950
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 2.44,
|
| 659 |
+
"learning_rate": 1.0416666666666667e-07,
|
| 660 |
+
"logps/chosen": -55.42496109008789,
|
| 661 |
+
"logps/rejected": -80.11758422851562,
|
| 662 |
+
"loss": 0.5346,
|
| 663 |
+
"losses/dpo": 0.5188722610473633,
|
| 664 |
+
"losses/sft": 1.9635555744171143,
|
| 665 |
+
"losses/total": 0.5188722610473633,
|
| 666 |
+
"ref_logps/chosen": -50.690673828125,
|
| 667 |
+
"ref_logps/rejected": -67.66686248779297,
|
| 668 |
+
"rewards/accuracies": 0.7055000066757202,
|
| 669 |
+
"rewards/chosen": -0.47342926263809204,
|
| 670 |
+
"rewards/margins": 0.7716432213783264,
|
| 671 |
+
"rewards/rejected": -1.2450724840164185,
|
| 672 |
+
"step": 975
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.5,
|
| 676 |
+
"learning_rate": 9.259259259259258e-08,
|
| 677 |
+
"logps/chosen": -55.2717399597168,
|
| 678 |
+
"logps/rejected": -78.23765563964844,
|
| 679 |
+
"loss": 0.5358,
|
| 680 |
+
"losses/dpo": 0.522473156452179,
|
| 681 |
+
"losses/sft": 1.9380919933319092,
|
| 682 |
+
"losses/total": 0.522473156452179,
|
| 683 |
+
"ref_logps/chosen": -50.576473236083984,
|
| 684 |
+
"ref_logps/rejected": -65.8857192993164,
|
| 685 |
+
"rewards/accuracies": 0.6880000233650208,
|
| 686 |
+
"rewards/chosen": -0.46952569484710693,
|
| 687 |
+
"rewards/margins": 0.7656675577163696,
|
| 688 |
+
"rewards/rejected": -1.2351933717727661,
|
| 689 |
+
"step": 1000
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 2.56,
|
| 693 |
+
"learning_rate": 8.101851851851852e-08,
|
| 694 |
+
"logps/chosen": -57.238101959228516,
|
| 695 |
+
"logps/rejected": -80.15925598144531,
|
| 696 |
+
"loss": 0.5477,
|
| 697 |
+
"losses/dpo": 0.5454632043838501,
|
| 698 |
+
"losses/sft": 1.9586001634597778,
|
| 699 |
+
"losses/total": 0.5454632043838501,
|
| 700 |
+
"ref_logps/chosen": -51.92245101928711,
|
| 701 |
+
"ref_logps/rejected": -67.11357879638672,
|
| 702 |
+
"rewards/accuracies": 0.6919999718666077,
|
| 703 |
+
"rewards/chosen": -0.5315641164779663,
|
| 704 |
+
"rewards/margins": 0.7730042934417725,
|
| 705 |
+
"rewards/rejected": -1.3045685291290283,
|
| 706 |
+
"step": 1025
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 2.62,
|
| 710 |
+
"learning_rate": 6.944444444444444e-08,
|
| 711 |
+
"logps/chosen": -56.340274810791016,
|
| 712 |
+
"logps/rejected": -78.74750518798828,
|
| 713 |
+
"loss": 0.5353,
|
| 714 |
+
"losses/dpo": 0.564401388168335,
|
| 715 |
+
"losses/sft": 1.9238648414611816,
|
| 716 |
+
"losses/total": 0.564401388168335,
|
| 717 |
+
"ref_logps/chosen": -51.298805236816406,
|
| 718 |
+
"ref_logps/rejected": -66.08538818359375,
|
| 719 |
+
"rewards/accuracies": 0.7080000042915344,
|
| 720 |
+
"rewards/chosen": -0.5041469931602478,
|
| 721 |
+
"rewards/margins": 0.7620646953582764,
|
| 722 |
+
"rewards/rejected": -1.2662116289138794,
|
| 723 |
+
"step": 1050
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 2.69,
|
| 727 |
+
"learning_rate": 5.787037037037037e-08,
|
| 728 |
+
"logps/chosen": -58.54380416870117,
|
| 729 |
+
"logps/rejected": -82.85121154785156,
|
| 730 |
+
"loss": 0.5379,
|
| 731 |
+
"losses/dpo": 0.4951947033405304,
|
| 732 |
+
"losses/sft": 1.927669882774353,
|
| 733 |
+
"losses/total": 0.4951947033405304,
|
| 734 |
+
"ref_logps/chosen": -53.125526428222656,
|
| 735 |
+
"ref_logps/rejected": -69.30874633789062,
|
| 736 |
+
"rewards/accuracies": 0.699999988079071,
|
| 737 |
+
"rewards/chosen": -0.5418279767036438,
|
| 738 |
+
"rewards/margins": 0.8124186992645264,
|
| 739 |
+
"rewards/rejected": -1.3542466163635254,
|
| 740 |
+
"step": 1075
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.75,
|
| 744 |
+
"learning_rate": 4.629629629629629e-08,
|
| 745 |
+
"logps/chosen": -56.923858642578125,
|
| 746 |
+
"logps/rejected": -81.64384460449219,
|
| 747 |
+
"loss": 0.5352,
|
| 748 |
+
"losses/dpo": 0.5503857135772705,
|
| 749 |
+
"losses/sft": 1.9750890731811523,
|
| 750 |
+
"losses/total": 0.5503857135772705,
|
| 751 |
+
"ref_logps/chosen": -51.61204528808594,
|
| 752 |
+
"ref_logps/rejected": -68.07569122314453,
|
| 753 |
+
"rewards/accuracies": 0.7145000696182251,
|
| 754 |
+
"rewards/chosen": -0.5311816930770874,
|
| 755 |
+
"rewards/margins": 0.8256343603134155,
|
| 756 |
+
"rewards/rejected": -1.356816053390503,
|
| 757 |
+
"step": 1100
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 2.81,
|
| 761 |
+
"learning_rate": 3.472222222222222e-08,
|
| 762 |
+
"logps/chosen": -57.147186279296875,
|
| 763 |
+
"logps/rejected": -79.592529296875,
|
| 764 |
+
"loss": 0.541,
|
| 765 |
+
"losses/dpo": 0.5187271237373352,
|
| 766 |
+
"losses/sft": 1.9453308582305908,
|
| 767 |
+
"losses/total": 0.5187271237373352,
|
| 768 |
+
"ref_logps/chosen": -52.0300178527832,
|
| 769 |
+
"ref_logps/rejected": -66.7751693725586,
|
| 770 |
+
"rewards/accuracies": 0.7010000348091125,
|
| 771 |
+
"rewards/chosen": -0.5117166042327881,
|
| 772 |
+
"rewards/margins": 0.7700192332267761,
|
| 773 |
+
"rewards/rejected": -1.281735897064209,
|
| 774 |
+
"step": 1125
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 2.88,
|
| 778 |
+
"learning_rate": 2.3148148148148144e-08,
|
| 779 |
+
"logps/chosen": -56.04924011230469,
|
| 780 |
+
"logps/rejected": -82.78221893310547,
|
| 781 |
+
"loss": 0.5138,
|
| 782 |
+
"losses/dpo": 0.5014563202857971,
|
| 783 |
+
"losses/sft": 1.892021656036377,
|
| 784 |
+
"losses/total": 0.5014563202857971,
|
| 785 |
+
"ref_logps/chosen": -50.947792053222656,
|
| 786 |
+
"ref_logps/rejected": -68.86062622070312,
|
| 787 |
+
"rewards/accuracies": 0.7239999175071716,
|
| 788 |
+
"rewards/chosen": -0.5101450085639954,
|
| 789 |
+
"rewards/margins": 0.8820143938064575,
|
| 790 |
+
"rewards/rejected": -1.3921594619750977,
|
| 791 |
+
"step": 1150
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 2.94,
|
| 795 |
+
"learning_rate": 1.1574074074074072e-08,
|
| 796 |
+
"logps/chosen": -57.95354461669922,
|
| 797 |
+
"logps/rejected": -81.48481750488281,
|
| 798 |
+
"loss": 0.5322,
|
| 799 |
+
"losses/dpo": 0.5276296138763428,
|
| 800 |
+
"losses/sft": 1.914094090461731,
|
| 801 |
+
"losses/total": 0.5276296138763428,
|
| 802 |
+
"ref_logps/chosen": -52.49882507324219,
|
| 803 |
+
"ref_logps/rejected": -68.10453796386719,
|
| 804 |
+
"rewards/accuracies": 0.7064999938011169,
|
| 805 |
+
"rewards/chosen": -0.5454717874526978,
|
| 806 |
+
"rewards/margins": 0.7925567030906677,
|
| 807 |
+
"rewards/rejected": -1.3380285501480103,
|
| 808 |
+
"step": 1175
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 3.0,
|
| 812 |
+
"learning_rate": 0.0,
|
| 813 |
+
"logps/chosen": -58.058658599853516,
|
| 814 |
+
"logps/rejected": -80.37024688720703,
|
| 815 |
+
"loss": 0.5418,
|
| 816 |
+
"losses/dpo": 0.5331112146377563,
|
| 817 |
+
"losses/sft": 1.9555615186691284,
|
| 818 |
+
"losses/total": 0.5331112146377563,
|
| 819 |
+
"ref_logps/chosen": -52.748592376708984,
|
| 820 |
+
"ref_logps/rejected": -67.33932495117188,
|
| 821 |
+
"rewards/accuracies": 0.6899999976158142,
|
| 822 |
+
"rewards/chosen": -0.5310066342353821,
|
| 823 |
+
"rewards/margins": 0.7720865607261658,
|
| 824 |
+
"rewards/rejected": -1.3030930757522583,
|
| 825 |
+
"step": 1200
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 3.0,
|
| 829 |
+
"step": 1200,
|
| 830 |
+
"total_flos": 0.0,
|
| 831 |
+
"train_loss": 0.5943309187889099,
|
| 832 |
+
"train_runtime": 16038.0463,
|
| 833 |
+
"train_samples_per_second": 5.984,
|
| 834 |
+
"train_steps_per_second": 0.075
|
| 835 |
+
}
|
| 836 |
+
],
|
| 837 |
+
"logging_steps": 25,
|
| 838 |
+
"max_steps": 1200,
|
| 839 |
+
"num_input_tokens_seen": 0,
|
| 840 |
+
"num_train_epochs": 3,
|
| 841 |
+
"save_steps": 500,
|
| 842 |
+
"stateful_callbacks": {},
|
| 843 |
+
"total_flos": 0.0,
|
| 844 |
+
"train_batch_size": 4,
|
| 845 |
+
"trial_name": null,
|
| 846 |
+
"trial_params": null
|
| 847 |
+
}
|