Commit
•
b179e6a
1
Parent(s):
ba4ab9d
Upload folder using huggingface_hub
Browse files- README.md +9 -0
- adapter_config.json +26 -0
- adapter_model.bin +3 -0
- config.json +45 -0
- non_lora_trainables.bin +3 -0
- trainer_state.json +655 -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 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"auto_mapping": null,
|
| 3 |
+
"base_model_name_or_path": "/home/ubuntu/LLaVA/checkpoints/llava-v1.5-13b-steer-lora-helpsteerformat-interleaf-lrv3067llava3067-likert-1222_interleavedgpt4v_responseonly-pure-merged",
|
| 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 |
+
"v_proj",
|
| 18 |
+
"down_proj",
|
| 19 |
+
"q_proj",
|
| 20 |
+
"o_proj",
|
| 21 |
+
"gate_proj",
|
| 22 |
+
"k_proj",
|
| 23 |
+
"up_proj"
|
| 24 |
+
],
|
| 25 |
+
"task_type": "CAUSAL_LM"
|
| 26 |
+
}
|
adapter_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d29226f1e5bc7cde54818e30abd0213f96e47a047e3fcd2361f415ea8c9be95
|
| 3 |
+
size 1001584333
|
config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/home/ubuntu/LLaVA/checkpoints/llava-v1.5-13b-steer-lora-helpsteerformat-interleaf-lrv3067llava3067-likert-1222_interleavedgpt4v_responseonly-pure-merged",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlavaLlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"eos_token_id": 2,
|
| 8 |
+
"freeze_mm_mlp_adapter": false,
|
| 9 |
+
"freeze_mm_vision_resampler": false,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 5120,
|
| 12 |
+
"image_aspect_ratio": "pad",
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 13824,
|
| 15 |
+
"max_length": 4096,
|
| 16 |
+
"max_position_embeddings": 4096,
|
| 17 |
+
"mm_hidden_size": 1024,
|
| 18 |
+
"mm_projector_lr": 2e-05,
|
| 19 |
+
"mm_projector_type": "mlp2x_gelu",
|
| 20 |
+
"mm_resampler_type": null,
|
| 21 |
+
"mm_use_im_patch_token": false,
|
| 22 |
+
"mm_use_im_start_end": false,
|
| 23 |
+
"mm_vision_select_feature": "patch",
|
| 24 |
+
"mm_vision_select_layer": -2,
|
| 25 |
+
"mm_vision_tower": "openai/clip-vit-large-patch14-336",
|
| 26 |
+
"model_type": "llava",
|
| 27 |
+
"num_attention_heads": 40,
|
| 28 |
+
"num_hidden_layers": 40,
|
| 29 |
+
"num_key_value_heads": 40,
|
| 30 |
+
"pad_token_id": 0,
|
| 31 |
+
"pretraining_tp": 1,
|
| 32 |
+
"rms_norm_eps": 1e-05,
|
| 33 |
+
"rope_scaling": null,
|
| 34 |
+
"tie_word_embeddings": false,
|
| 35 |
+
"tokenizer_model_max_length": 2048,
|
| 36 |
+
"tokenizer_padding_side": "right",
|
| 37 |
+
"torch_dtype": "float16",
|
| 38 |
+
"transformers_version": "4.31.0",
|
| 39 |
+
"tune_mm_mlp_adapter": false,
|
| 40 |
+
"tune_mm_vision_resampler": false,
|
| 41 |
+
"unfreeze_mm_vision_tower": false,
|
| 42 |
+
"use_cache": true,
|
| 43 |
+
"use_mm_proj": true,
|
| 44 |
+
"vocab_size": 32000
|
| 45 |
+
}
|
non_lora_trainables.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d08a173a526e8a979404fc833be6312a1638363d2545270d93170a3e3ebf41c8
|
| 3 |
+
size 62936807
|
trainer_state.json
ADDED
|
@@ -0,0 +1,655 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.8,
|
| 5 |
+
"global_step": 21,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0,
|
| 12 |
+
"logps_train/chosen": -100.49485778808594,
|
| 13 |
+
"logps_train/ref_chosen": -100.5,
|
| 14 |
+
"logps_train/ref_rejected": -105.0,
|
| 15 |
+
"logps_train/rejected": -104.80752563476562,
|
| 16 |
+
"rewards_train/accuracies": 0.46875,
|
| 17 |
+
"rewards_train/chosen": -0.020498279482126236,
|
| 18 |
+
"rewards_train/margins": -0.021032828837633133,
|
| 19 |
+
"rewards_train/rejected": 0.000534549355506897,
|
| 20 |
+
"step": 0
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0,
|
| 24 |
+
"logps_train/chosen": -89.90950012207031,
|
| 25 |
+
"logps_train/ref_chosen": -90.0,
|
| 26 |
+
"logps_train/ref_rejected": -101.0,
|
| 27 |
+
"logps_train/rejected": -100.86872863769531,
|
| 28 |
+
"rewards_train/accuracies": 0.5,
|
| 29 |
+
"rewards_train/chosen": 0.0191262885928154,
|
| 30 |
+
"rewards_train/margins": 0.0020734891295433044,
|
| 31 |
+
"rewards_train/rejected": 0.017052799463272095,
|
| 32 |
+
"step": 0
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.13,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 0.7043,
|
| 38 |
+
"step": 1
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.13,
|
| 42 |
+
"logps_train/chosen": -104.38166809082031,
|
| 43 |
+
"logps_train/ref_chosen": -104.5,
|
| 44 |
+
"logps_train/ref_rejected": -98.0,
|
| 45 |
+
"logps_train/rejected": -97.89907836914062,
|
| 46 |
+
"rewards_train/accuracies": 0.515625,
|
| 47 |
+
"rewards_train/chosen": 0.03059956058859825,
|
| 48 |
+
"rewards_train/margins": 0.004916800186038017,
|
| 49 |
+
"rewards_train/rejected": 0.025682760402560234,
|
| 50 |
+
"step": 1
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.13,
|
| 54 |
+
"logps_train/chosen": -94.88069152832031,
|
| 55 |
+
"logps_train/ref_chosen": -95.0,
|
| 56 |
+
"logps_train/ref_rejected": -95.0,
|
| 57 |
+
"logps_train/rejected": -95.29293823242188,
|
| 58 |
+
"rewards_train/accuracies": 0.546875,
|
| 59 |
+
"rewards_train/chosen": -0.009438544511795044,
|
| 60 |
+
"rewards_train/margins": -0.020688317716121674,
|
| 61 |
+
"rewards_train/rejected": 0.01124977320432663,
|
| 62 |
+
"step": 1
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.27,
|
| 66 |
+
"learning_rate": 4.9692208514878444e-05,
|
| 67 |
+
"loss": 0.7061,
|
| 68 |
+
"step": 2
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.27,
|
| 72 |
+
"logps_train/chosen": -82.19304656982422,
|
| 73 |
+
"logps_train/ref_chosen": -82.5,
|
| 74 |
+
"logps_train/ref_rejected": -89.0,
|
| 75 |
+
"logps_train/rejected": -88.60076904296875,
|
| 76 |
+
"rewards_train/accuracies": 0.40625,
|
| 77 |
+
"rewards_train/chosen": 0.1768513023853302,
|
| 78 |
+
"rewards_train/margins": -0.027216821908950806,
|
| 79 |
+
"rewards_train/rejected": 0.204068124294281,
|
| 80 |
+
"step": 2
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.27,
|
| 84 |
+
"logps_train/chosen": -99.76242065429688,
|
| 85 |
+
"logps_train/ref_chosen": -100.0,
|
| 86 |
+
"logps_train/ref_rejected": -106.0,
|
| 87 |
+
"logps_train/rejected": -105.78941345214844,
|
| 88 |
+
"rewards_train/accuracies": 0.609375,
|
| 89 |
+
"rewards_train/chosen": 0.19887080788612366,
|
| 90 |
+
"rewards_train/margins": 0.0869825929403305,
|
| 91 |
+
"rewards_train/rejected": 0.11188821494579315,
|
| 92 |
+
"step": 2
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.4,
|
| 96 |
+
"learning_rate": 4.877641290737884e-05,
|
| 97 |
+
"loss": 0.6893,
|
| 98 |
+
"step": 3
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.4,
|
| 102 |
+
"logps_train/chosen": -85.4503173828125,
|
| 103 |
+
"logps_train/ref_chosen": -86.0,
|
| 104 |
+
"logps_train/ref_rejected": -90.0,
|
| 105 |
+
"logps_train/rejected": -89.7308120727539,
|
| 106 |
+
"rewards_train/accuracies": 0.5625,
|
| 107 |
+
"rewards_train/chosen": 0.1976955533027649,
|
| 108 |
+
"rewards_train/margins": 0.044912755489349365,
|
| 109 |
+
"rewards_train/rejected": 0.15278279781341553,
|
| 110 |
+
"step": 3
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.4,
|
| 114 |
+
"logps_train/chosen": -91.95987701416016,
|
| 115 |
+
"logps_train/ref_chosen": -92.0,
|
| 116 |
+
"logps_train/ref_rejected": -77.5,
|
| 117 |
+
"logps_train/rejected": -77.18235778808594,
|
| 118 |
+
"rewards_train/accuracies": 0.484375,
|
| 119 |
+
"rewards_train/chosen": 0.17484739422798157,
|
| 120 |
+
"rewards_train/margins": -0.023889049887657166,
|
| 121 |
+
"rewards_train/rejected": 0.19873644411563873,
|
| 122 |
+
"step": 3
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.53,
|
| 126 |
+
"learning_rate": 4.72751631047092e-05,
|
| 127 |
+
"loss": 0.7254,
|
| 128 |
+
"step": 4
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.53,
|
| 132 |
+
"logps_train/chosen": -88.06124114990234,
|
| 133 |
+
"logps_train/ref_chosen": -88.5,
|
| 134 |
+
"logps_train/ref_rejected": -91.5,
|
| 135 |
+
"logps_train/rejected": -91.30616760253906,
|
| 136 |
+
"rewards_train/accuracies": 0.5625,
|
| 137 |
+
"rewards_train/chosen": 0.2691383361816406,
|
| 138 |
+
"rewards_train/margins": 0.15871848165988922,
|
| 139 |
+
"rewards_train/rejected": 0.1104198545217514,
|
| 140 |
+
"step": 4
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.53,
|
| 144 |
+
"logps_train/chosen": -104.14408111572266,
|
| 145 |
+
"logps_train/ref_chosen": -105.0,
|
| 146 |
+
"logps_train/ref_rejected": -104.5,
|
| 147 |
+
"logps_train/rejected": -103.88468933105469,
|
| 148 |
+
"rewards_train/accuracies": 0.578125,
|
| 149 |
+
"rewards_train/chosen": 0.3913380205631256,
|
| 150 |
+
"rewards_train/margins": 0.13361021876335144,
|
| 151 |
+
"rewards_train/rejected": 0.25772780179977417,
|
| 152 |
+
"step": 4
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.67,
|
| 156 |
+
"learning_rate": 4.522542485937369e-05,
|
| 157 |
+
"loss": 0.672,
|
| 158 |
+
"step": 5
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.67,
|
| 162 |
+
"logps_train/chosen": -108.3313980102539,
|
| 163 |
+
"logps_train/ref_chosen": -109.0,
|
| 164 |
+
"logps_train/ref_rejected": -99.5,
|
| 165 |
+
"logps_train/rejected": -98.8963623046875,
|
| 166 |
+
"rewards_train/accuracies": 0.640625,
|
| 167 |
+
"rewards_train/chosen": 0.48469099402427673,
|
| 168 |
+
"rewards_train/margins": 0.2424430549144745,
|
| 169 |
+
"rewards_train/rejected": 0.24224793910980225,
|
| 170 |
+
"step": 5
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.67,
|
| 174 |
+
"logps_train/chosen": -81.63034057617188,
|
| 175 |
+
"logps_train/ref_chosen": -82.5,
|
| 176 |
+
"logps_train/ref_rejected": -80.0,
|
| 177 |
+
"logps_train/rejected": -78.99322509765625,
|
| 178 |
+
"rewards_train/accuracies": 0.484375,
|
| 179 |
+
"rewards_train/chosen": 0.48524150252342224,
|
| 180 |
+
"rewards_train/margins": 0.020122677087783813,
|
| 181 |
+
"rewards_train/rejected": 0.4651188254356384,
|
| 182 |
+
"step": 5
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.8,
|
| 186 |
+
"learning_rate": 4.267766952966369e-05,
|
| 187 |
+
"loss": 0.7247,
|
| 188 |
+
"step": 6
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.8,
|
| 192 |
+
"logps_train/chosen": -91.17040252685547,
|
| 193 |
+
"logps_train/ref_chosen": -92.0,
|
| 194 |
+
"logps_train/ref_rejected": -89.5,
|
| 195 |
+
"logps_train/rejected": -88.93339538574219,
|
| 196 |
+
"rewards_train/accuracies": 0.5,
|
| 197 |
+
"rewards_train/chosen": 0.522466778755188,
|
| 198 |
+
"rewards_train/margins": 0.16933834552764893,
|
| 199 |
+
"rewards_train/rejected": 0.35312843322753906,
|
| 200 |
+
"step": 6
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.8,
|
| 204 |
+
"logps_train/chosen": -87.31026458740234,
|
| 205 |
+
"logps_train/ref_chosen": -88.5,
|
| 206 |
+
"logps_train/ref_rejected": -94.0,
|
| 207 |
+
"logps_train/rejected": -93.22328186035156,
|
| 208 |
+
"rewards_train/accuracies": 0.625,
|
| 209 |
+
"rewards_train/chosen": 0.4981892704963684,
|
| 210 |
+
"rewards_train/margins": 0.22706186771392822,
|
| 211 |
+
"rewards_train/rejected": 0.2711274027824402,
|
| 212 |
+
"step": 6
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.93,
|
| 216 |
+
"learning_rate": 3.969463130731183e-05,
|
| 217 |
+
"loss": 0.6977,
|
| 218 |
+
"step": 7
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.93,
|
| 222 |
+
"logps_train/chosen": -97.33760833740234,
|
| 223 |
+
"logps_train/ref_chosen": -107.0,
|
| 224 |
+
"logps_train/ref_rejected": -104.0,
|
| 225 |
+
"logps_train/rejected": -112.84634399414062,
|
| 226 |
+
"rewards_train/accuracies": 0.921875,
|
| 227 |
+
"rewards_train/chosen": 4.983658790588379,
|
| 228 |
+
"rewards_train/margins": 9.4771409034729,
|
| 229 |
+
"rewards_train/rejected": -4.4934821128845215,
|
| 230 |
+
"step": 7
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.93,
|
| 234 |
+
"logps_train/chosen": -87.83363342285156,
|
| 235 |
+
"logps_train/ref_chosen": -96.0,
|
| 236 |
+
"logps_train/ref_rejected": -90.5,
|
| 237 |
+
"logps_train/rejected": -96.15026092529297,
|
| 238 |
+
"rewards_train/accuracies": 0.96875,
|
| 239 |
+
"rewards_train/chosen": 4.031304359436035,
|
| 240 |
+
"rewards_train/margins": 6.9581804275512695,
|
| 241 |
+
"rewards_train/rejected": -2.9268760681152344,
|
| 242 |
+
"step": 7
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.07,
|
| 246 |
+
"learning_rate": 3.634976249348867e-05,
|
| 247 |
+
"loss": 0.1455,
|
| 248 |
+
"step": 8
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.07,
|
| 252 |
+
"logps_train/chosen": -92.92860412597656,
|
| 253 |
+
"logps_train/ref_chosen": -104.0,
|
| 254 |
+
"logps_train/ref_rejected": -91.0,
|
| 255 |
+
"logps_train/rejected": -100.07354736328125,
|
| 256 |
+
"rewards_train/accuracies": 1.0,
|
| 257 |
+
"rewards_train/chosen": 5.565237522125244,
|
| 258 |
+
"rewards_train/margins": 10.037434577941895,
|
| 259 |
+
"rewards_train/rejected": -4.47219705581665,
|
| 260 |
+
"step": 8
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.07,
|
| 264 |
+
"logps_train/chosen": -81.51614379882812,
|
| 265 |
+
"logps_train/ref_chosen": -90.0,
|
| 266 |
+
"logps_train/ref_rejected": -91.0,
|
| 267 |
+
"logps_train/rejected": -98.16658782958984,
|
| 268 |
+
"rewards_train/accuracies": 1.0,
|
| 269 |
+
"rewards_train/chosen": 4.322493553161621,
|
| 270 |
+
"rewards_train/margins": 7.822535753250122,
|
| 271 |
+
"rewards_train/rejected": -3.500042200088501,
|
| 272 |
+
"step": 8
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.2,
|
| 276 |
+
"learning_rate": 3.272542485937369e-05,
|
| 277 |
+
"loss": 0.0635,
|
| 278 |
+
"step": 9
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.2,
|
| 282 |
+
"logps_train/chosen": -89.58404541015625,
|
| 283 |
+
"logps_train/ref_chosen": -100.0,
|
| 284 |
+
"logps_train/ref_rejected": -91.5,
|
| 285 |
+
"logps_train/rejected": -101.0616455078125,
|
| 286 |
+
"rewards_train/accuracies": 0.984375,
|
| 287 |
+
"rewards_train/chosen": 5.313934326171875,
|
| 288 |
+
"rewards_train/margins": 10.12344217300415,
|
| 289 |
+
"rewards_train/rejected": -4.809507846832275,
|
| 290 |
+
"step": 9
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.2,
|
| 294 |
+
"logps_train/chosen": -87.73910522460938,
|
| 295 |
+
"logps_train/ref_chosen": -98.0,
|
| 296 |
+
"logps_train/ref_rejected": -95.5,
|
| 297 |
+
"logps_train/rejected": -103.09068298339844,
|
| 298 |
+
"rewards_train/accuracies": 1.0,
|
| 299 |
+
"rewards_train/chosen": 5.000567436218262,
|
| 300 |
+
"rewards_train/margins": 8.863048553466797,
|
| 301 |
+
"rewards_train/rejected": -3.862481117248535,
|
| 302 |
+
"step": 9
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.33,
|
| 306 |
+
"learning_rate": 2.8910861626005776e-05,
|
| 307 |
+
"loss": 0.0449,
|
| 308 |
+
"step": 10
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.33,
|
| 312 |
+
"logps_train/chosen": -90.21186065673828,
|
| 313 |
+
"logps_train/ref_chosen": -99.5,
|
| 314 |
+
"logps_train/ref_rejected": -97.0,
|
| 315 |
+
"logps_train/rejected": -104.75820922851562,
|
| 316 |
+
"rewards_train/accuracies": 1.0,
|
| 317 |
+
"rewards_train/chosen": 4.69606876373291,
|
| 318 |
+
"rewards_train/margins": 8.679665327072144,
|
| 319 |
+
"rewards_train/rejected": -3.9835965633392334,
|
| 320 |
+
"step": 10
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.33,
|
| 324 |
+
"logps_train/chosen": -81.06768798828125,
|
| 325 |
+
"logps_train/ref_chosen": -91.0,
|
| 326 |
+
"logps_train/ref_rejected": -97.5,
|
| 327 |
+
"logps_train/rejected": -106.78590393066406,
|
| 328 |
+
"rewards_train/accuracies": 0.984375,
|
| 329 |
+
"rewards_train/chosen": 5.148348808288574,
|
| 330 |
+
"rewards_train/margins": 9.690350532531738,
|
| 331 |
+
"rewards_train/rejected": -4.542001724243164,
|
| 332 |
+
"step": 10
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.47,
|
| 336 |
+
"learning_rate": 2.5e-05,
|
| 337 |
+
"loss": 0.0452,
|
| 338 |
+
"step": 11
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.47,
|
| 342 |
+
"logps_train/chosen": -85.45967102050781,
|
| 343 |
+
"logps_train/ref_chosen": -97.0,
|
| 344 |
+
"logps_train/ref_rejected": -99.0,
|
| 345 |
+
"logps_train/rejected": -109.41343688964844,
|
| 346 |
+
"rewards_train/accuracies": 1.0,
|
| 347 |
+
"rewards_train/chosen": 5.63247013092041,
|
| 348 |
+
"rewards_train/margins": 10.989093780517578,
|
| 349 |
+
"rewards_train/rejected": -5.356623649597168,
|
| 350 |
+
"step": 11
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.47,
|
| 354 |
+
"logps_train/chosen": -71.80900573730469,
|
| 355 |
+
"logps_train/ref_chosen": -81.0,
|
| 356 |
+
"logps_train/ref_rejected": -90.0,
|
| 357 |
+
"logps_train/rejected": -97.481689453125,
|
| 358 |
+
"rewards_train/accuracies": 0.96875,
|
| 359 |
+
"rewards_train/chosen": 4.653310775756836,
|
| 360 |
+
"rewards_train/margins": 8.362460613250732,
|
| 361 |
+
"rewards_train/rejected": -3.7091498374938965,
|
| 362 |
+
"step": 11
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.6,
|
| 366 |
+
"learning_rate": 2.1089138373994223e-05,
|
| 367 |
+
"loss": 0.0475,
|
| 368 |
+
"step": 12
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.6,
|
| 372 |
+
"logps_train/chosen": -79.43351745605469,
|
| 373 |
+
"logps_train/ref_chosen": -90.0,
|
| 374 |
+
"logps_train/ref_rejected": -96.0,
|
| 375 |
+
"logps_train/rejected": -105.02793884277344,
|
| 376 |
+
"rewards_train/accuracies": 1.0,
|
| 377 |
+
"rewards_train/chosen": 5.266395568847656,
|
| 378 |
+
"rewards_train/margins": 9.863905906677246,
|
| 379 |
+
"rewards_train/rejected": -4.59751033782959,
|
| 380 |
+
"step": 12
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.6,
|
| 384 |
+
"logps_train/chosen": -73.93589782714844,
|
| 385 |
+
"logps_train/ref_chosen": -83.0,
|
| 386 |
+
"logps_train/ref_rejected": -85.0,
|
| 387 |
+
"logps_train/rejected": -93.78298950195312,
|
| 388 |
+
"rewards_train/accuracies": 1.0,
|
| 389 |
+
"rewards_train/chosen": 4.597479343414307,
|
| 390 |
+
"rewards_train/margins": 8.91060733795166,
|
| 391 |
+
"rewards_train/rejected": -4.3131279945373535,
|
| 392 |
+
"step": 12
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.73,
|
| 396 |
+
"learning_rate": 1.7274575140626318e-05,
|
| 397 |
+
"loss": 0.0501,
|
| 398 |
+
"step": 13
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.73,
|
| 402 |
+
"logps_train/chosen": -90.48731231689453,
|
| 403 |
+
"logps_train/ref_chosen": -105.0,
|
| 404 |
+
"logps_train/ref_rejected": -101.0,
|
| 405 |
+
"logps_train/rejected": -113.04267883300781,
|
| 406 |
+
"rewards_train/accuracies": 1.0,
|
| 407 |
+
"rewards_train/chosen": 7.302729606628418,
|
| 408 |
+
"rewards_train/margins": 13.378514289855957,
|
| 409 |
+
"rewards_train/rejected": -6.075784683227539,
|
| 410 |
+
"step": 13
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.73,
|
| 414 |
+
"logps_train/chosen": -78.76904296875,
|
| 415 |
+
"logps_train/ref_chosen": -91.0,
|
| 416 |
+
"logps_train/ref_rejected": -85.0,
|
| 417 |
+
"logps_train/rejected": -94.63651275634766,
|
| 418 |
+
"rewards_train/accuracies": 1.0,
|
| 419 |
+
"rewards_train/chosen": 5.878904819488525,
|
| 420 |
+
"rewards_train/margins": 10.652726650238037,
|
| 421 |
+
"rewards_train/rejected": -4.773821830749512,
|
| 422 |
+
"step": 13
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.87,
|
| 426 |
+
"learning_rate": 1.3650237506511331e-05,
|
| 427 |
+
"loss": 0.023,
|
| 428 |
+
"step": 14
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.87,
|
| 432 |
+
"logps_train/chosen": -82.4466552734375,
|
| 433 |
+
"logps_train/ref_chosen": -93.0,
|
| 434 |
+
"logps_train/ref_rejected": -108.5,
|
| 435 |
+
"logps_train/rejected": -120.84589385986328,
|
| 436 |
+
"rewards_train/accuracies": 1.0,
|
| 437 |
+
"rewards_train/chosen": 5.197817802429199,
|
| 438 |
+
"rewards_train/margins": 11.340189933776855,
|
| 439 |
+
"rewards_train/rejected": -6.142372131347656,
|
| 440 |
+
"step": 14
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.87,
|
| 444 |
+
"logps_train/chosen": -81.03172302246094,
|
| 445 |
+
"logps_train/ref_chosen": -94.0,
|
| 446 |
+
"logps_train/ref_rejected": -93.0,
|
| 447 |
+
"logps_train/rejected": -104.14176940917969,
|
| 448 |
+
"rewards_train/accuracies": 1.0,
|
| 449 |
+
"rewards_train/chosen": 6.500617027282715,
|
| 450 |
+
"rewards_train/margins": 12.0755934715271,
|
| 451 |
+
"rewards_train/rejected": -5.574976444244385,
|
| 452 |
+
"step": 14
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.0,
|
| 456 |
+
"learning_rate": 1.0305368692688174e-05,
|
| 457 |
+
"loss": 0.0299,
|
| 458 |
+
"step": 15
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.0,
|
| 462 |
+
"logps_train/chosen": -84.29070281982422,
|
| 463 |
+
"logps_train/ref_chosen": -97.5,
|
| 464 |
+
"logps_train/ref_rejected": -96.5,
|
| 465 |
+
"logps_train/rejected": -108.1887435913086,
|
| 466 |
+
"rewards_train/accuracies": 1.0,
|
| 467 |
+
"rewards_train/chosen": 6.537390232086182,
|
| 468 |
+
"rewards_train/margins": 12.432237148284912,
|
| 469 |
+
"rewards_train/rejected": -5.8948469161987305,
|
| 470 |
+
"step": 15
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.0,
|
| 474 |
+
"logps_train/chosen": -80.9233627319336,
|
| 475 |
+
"logps_train/ref_chosen": -93.5,
|
| 476 |
+
"logps_train/ref_rejected": -106.0,
|
| 477 |
+
"logps_train/rejected": -118.09178161621094,
|
| 478 |
+
"rewards_train/accuracies": 1.0,
|
| 479 |
+
"rewards_train/chosen": 6.2991838455200195,
|
| 480 |
+
"rewards_train/margins": 12.446146488189697,
|
| 481 |
+
"rewards_train/rejected": -6.146962642669678,
|
| 482 |
+
"step": 15
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.13,
|
| 486 |
+
"learning_rate": 7.3223304703363135e-06,
|
| 487 |
+
"loss": 0.028,
|
| 488 |
+
"step": 16
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.13,
|
| 492 |
+
"logps_train/chosen": -83.49127960205078,
|
| 493 |
+
"logps_train/ref_chosen": -95.0,
|
| 494 |
+
"logps_train/ref_rejected": -89.0,
|
| 495 |
+
"logps_train/rejected": -99.20130157470703,
|
| 496 |
+
"rewards_train/accuracies": 1.0,
|
| 497 |
+
"rewards_train/chosen": 5.888392448425293,
|
| 498 |
+
"rewards_train/margins": 11.02004623413086,
|
| 499 |
+
"rewards_train/rejected": -5.131653785705566,
|
| 500 |
+
"step": 16
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.13,
|
| 504 |
+
"logps_train/chosen": -82.60940551757812,
|
| 505 |
+
"logps_train/ref_chosen": -94.0,
|
| 506 |
+
"logps_train/ref_rejected": -88.5,
|
| 507 |
+
"logps_train/rejected": -98.06349182128906,
|
| 508 |
+
"rewards_train/accuracies": 1.0,
|
| 509 |
+
"rewards_train/chosen": 5.770737648010254,
|
| 510 |
+
"rewards_train/margins": 10.54662036895752,
|
| 511 |
+
"rewards_train/rejected": -4.775882720947266,
|
| 512 |
+
"step": 16
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 2.27,
|
| 516 |
+
"learning_rate": 4.7745751406263165e-06,
|
| 517 |
+
"loss": 0.034,
|
| 518 |
+
"step": 17
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 2.27,
|
| 522 |
+
"logps_train/chosen": -82.3560791015625,
|
| 523 |
+
"logps_train/ref_chosen": -96.0,
|
| 524 |
+
"logps_train/ref_rejected": -96.0,
|
| 525 |
+
"logps_train/rejected": -108.4398193359375,
|
| 526 |
+
"rewards_train/accuracies": 1.0,
|
| 527 |
+
"rewards_train/chosen": 6.842710018157959,
|
| 528 |
+
"rewards_train/margins": 13.140255451202393,
|
| 529 |
+
"rewards_train/rejected": -6.297545433044434,
|
| 530 |
+
"step": 17
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.27,
|
| 534 |
+
"logps_train/chosen": -79.5767822265625,
|
| 535 |
+
"logps_train/ref_chosen": -91.5,
|
| 536 |
+
"logps_train/ref_rejected": -100.0,
|
| 537 |
+
"logps_train/rejected": -114.11109924316406,
|
| 538 |
+
"rewards_train/accuracies": 1.0,
|
| 539 |
+
"rewards_train/chosen": 6.006286144256592,
|
| 540 |
+
"rewards_train/margins": 13.076303482055664,
|
| 541 |
+
"rewards_train/rejected": -7.070017337799072,
|
| 542 |
+
"step": 17
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.4,
|
| 546 |
+
"learning_rate": 2.7248368952908053e-06,
|
| 547 |
+
"loss": 0.0185,
|
| 548 |
+
"step": 18
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.4,
|
| 552 |
+
"logps_train/chosen": -87.6229019165039,
|
| 553 |
+
"logps_train/ref_chosen": -102.0,
|
| 554 |
+
"logps_train/ref_rejected": -99.0,
|
| 555 |
+
"logps_train/rejected": -112.02855682373047,
|
| 556 |
+
"rewards_train/accuracies": 1.0,
|
| 557 |
+
"rewards_train/chosen": 7.329846382141113,
|
| 558 |
+
"rewards_train/margins": 13.926156997680664,
|
| 559 |
+
"rewards_train/rejected": -6.596310615539551,
|
| 560 |
+
"step": 18
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.4,
|
| 564 |
+
"logps_train/chosen": -87.40618896484375,
|
| 565 |
+
"logps_train/ref_chosen": -101.5,
|
| 566 |
+
"logps_train/ref_rejected": -101.0,
|
| 567 |
+
"logps_train/rejected": -114.67952728271484,
|
| 568 |
+
"rewards_train/accuracies": 1.0,
|
| 569 |
+
"rewards_train/chosen": 7.131863594055176,
|
| 570 |
+
"rewards_train/margins": 13.977363586425781,
|
| 571 |
+
"rewards_train/rejected": -6.8454999923706055,
|
| 572 |
+
"step": 18
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 2.53,
|
| 576 |
+
"learning_rate": 1.2235870926211619e-06,
|
| 577 |
+
"loss": 0.0233,
|
| 578 |
+
"step": 19
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.53,
|
| 582 |
+
"logps_train/chosen": -81.33338928222656,
|
| 583 |
+
"logps_train/ref_chosen": -94.5,
|
| 584 |
+
"logps_train/ref_rejected": -101.0,
|
| 585 |
+
"logps_train/rejected": -115.27784729003906,
|
| 586 |
+
"rewards_train/accuracies": 1.0,
|
| 587 |
+
"rewards_train/chosen": 6.613577842712402,
|
| 588 |
+
"rewards_train/margins": 13.59971284866333,
|
| 589 |
+
"rewards_train/rejected": -6.986135005950928,
|
| 590 |
+
"step": 19
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.53,
|
| 594 |
+
"logps_train/chosen": -89.27815246582031,
|
| 595 |
+
"logps_train/ref_chosen": -101.0,
|
| 596 |
+
"logps_train/ref_rejected": -95.5,
|
| 597 |
+
"logps_train/rejected": -105.24420928955078,
|
| 598 |
+
"rewards_train/accuracies": 1.0,
|
| 599 |
+
"rewards_train/chosen": 5.84676456451416,
|
| 600 |
+
"rewards_train/margins": 10.66076374053955,
|
| 601 |
+
"rewards_train/rejected": -4.813999176025391,
|
| 602 |
+
"step": 19
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.67,
|
| 606 |
+
"learning_rate": 3.077914851215585e-07,
|
| 607 |
+
"loss": 0.0211,
|
| 608 |
+
"step": 20
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.67,
|
| 612 |
+
"logps_train/chosen": -66.01316833496094,
|
| 613 |
+
"logps_train/ref_chosen": -76.0,
|
| 614 |
+
"logps_train/ref_rejected": -78.0,
|
| 615 |
+
"logps_train/rejected": -87.24235534667969,
|
| 616 |
+
"rewards_train/accuracies": 1.0,
|
| 617 |
+
"rewards_train/chosen": 5.09859561920166,
|
| 618 |
+
"rewards_train/margins": 9.691985130310059,
|
| 619 |
+
"rewards_train/rejected": -4.593389511108398,
|
| 620 |
+
"step": 20
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.67,
|
| 624 |
+
"logps_train/chosen": -82.14631652832031,
|
| 625 |
+
"logps_train/ref_chosen": -96.0,
|
| 626 |
+
"logps_train/ref_rejected": -94.0,
|
| 627 |
+
"logps_train/rejected": -106.15888214111328,
|
| 628 |
+
"rewards_train/accuracies": 1.0,
|
| 629 |
+
"rewards_train/chosen": 6.794517993927002,
|
| 630 |
+
"rewards_train/margins": 12.825128555297852,
|
| 631 |
+
"rewards_train/rejected": -6.03061056137085,
|
| 632 |
+
"step": 20
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.8,
|
| 636 |
+
"learning_rate": 0.0,
|
| 637 |
+
"loss": 0.0335,
|
| 638 |
+
"step": 21
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.8,
|
| 642 |
+
"step": 21,
|
| 643 |
+
"total_flos": 0.0,
|
| 644 |
+
"train_loss": 0.2632110191597825,
|
| 645 |
+
"train_runtime": 244.5091,
|
| 646 |
+
"train_samples_per_second": 11.153,
|
| 647 |
+
"train_steps_per_second": 0.086
|
| 648 |
+
}
|
| 649 |
+
],
|
| 650 |
+
"max_steps": 21,
|
| 651 |
+
"num_train_epochs": 3,
|
| 652 |
+
"total_flos": 0.0,
|
| 653 |
+
"trial_name": null,
|
| 654 |
+
"trial_params": null
|
| 655 |
+
}
|