Model save
Browse files- README.md +68 -0
- adapter_config.json +25 -0
- adapter_model.safetensors +3 -0
- all_results.json +22 -0
- eval_results.json +16 -0
- runs/Jan05_23-27-01_amaterasu/events.out.tfevents.1704497299.amaterasu.1124388.1 +3 -0
- runs/Jan05_23-32-55_amaterasu/events.out.tfevents.1704497659.amaterasu.1124388.2 +3 -0
- runs/Jan05_23-39-40_amaterasu/events.out.tfevents.1704498064.amaterasu.1124388.3 +3 -0
- runs/Jan05_23-47-10_amaterasu/events.out.tfevents.1704498636.amaterasu.1197664.0 +3 -0
- runs/Jan05_23-47-10_amaterasu/events.out.tfevents.1704499383.amaterasu.1197664.1 +3 -0
- runs/Jan06_00-04-38_amaterasu/events.out.tfevents.1704499547.amaterasu.1204821.0 +3 -0
- runs/Jan06_00-08-44_amaterasu/events.out.tfevents.1704499811.amaterasu.1206551.0 +3 -0
- runs/Jan06_00-12-03_amaterasu/events.out.tfevents.1704500027.amaterasu.1207814.0 +3 -0
- runs/Jan06_00-12-03_amaterasu/events.out.tfevents.1704594215.amaterasu.1207814.1 +3 -0
- special_tokens_map.json +30 -0
- tokenizer.json +0 -0
- tokenizer_config.json +41 -0
- train_results.json +9 -0
- trainer_state.json +2116 -0
- training_args.bin +3 -0
README.md
ADDED
|
@@ -0,0 +1,68 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: mistralai/Mistral-7B-v0.1
|
| 4 |
+
tags:
|
| 5 |
+
- generated_from_trainer
|
| 6 |
+
model-index:
|
| 7 |
+
- name: zephyr-7b-dpo-lora
|
| 8 |
+
results: []
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 12 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 13 |
+
|
| 14 |
+
# zephyr-7b-dpo-lora
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of [mistralai/Mistral-7B-v0.1](https://huggingface.co/mistralai/Mistral-7B-v0.1) on an unknown dataset.
|
| 17 |
+
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 0.2082
|
| 19 |
+
- Rewards/chosen: 1.3857
|
| 20 |
+
- Rewards/rejected: -0.9066
|
| 21 |
+
- Rewards/accuracies: 0.9414
|
| 22 |
+
- Rewards/margins: 2.2923
|
| 23 |
+
- Logps/rejected: -388.5903
|
| 24 |
+
- Logps/chosen: -238.5479
|
| 25 |
+
- Logits/rejected: -2.7219
|
| 26 |
+
- Logits/chosen: -2.6178
|
| 27 |
+
|
| 28 |
+
## Model description
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Intended uses & limitations
|
| 33 |
+
|
| 34 |
+
More information needed
|
| 35 |
+
|
| 36 |
+
## Training and evaluation data
|
| 37 |
+
|
| 38 |
+
More information needed
|
| 39 |
+
|
| 40 |
+
## Training procedure
|
| 41 |
+
|
| 42 |
+
### Training hyperparameters
|
| 43 |
+
|
| 44 |
+
The following hyperparameters were used during training:
|
| 45 |
+
- learning_rate: 5e-07
|
| 46 |
+
- train_batch_size: 4
|
| 47 |
+
- eval_batch_size: 8
|
| 48 |
+
- seed: 42
|
| 49 |
+
- gradient_accumulation_steps: 32
|
| 50 |
+
- total_train_batch_size: 128
|
| 51 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 52 |
+
- lr_scheduler_type: linear
|
| 53 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 54 |
+
- num_epochs: 1
|
| 55 |
+
|
| 56 |
+
### Training results
|
| 57 |
+
|
| 58 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
| 59 |
+
|:-------------:|:-----:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
| 60 |
+
| 0.2019 | 1.0 | 1470 | 0.2082 | 1.3857 | -0.9066 | 0.9414 | 2.2923 | -388.5903 | -238.5479 | -2.7219 | -2.6178 |
|
| 61 |
+
|
| 62 |
+
|
| 63 |
+
### Framework versions
|
| 64 |
+
|
| 65 |
+
- Transformers 4.35.0
|
| 66 |
+
- Pytorch 2.1.1+cu121
|
| 67 |
+
- Datasets 2.14.6
|
| 68 |
+
- Tokenizers 0.14.1
|
adapter_config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "mistralai/Mistral-7B-v0.1",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"lora_alpha": 16,
|
| 12 |
+
"lora_dropout": 0.1,
|
| 13 |
+
"modules_to_save": null,
|
| 14 |
+
"peft_type": "LORA",
|
| 15 |
+
"r": 64,
|
| 16 |
+
"rank_pattern": {},
|
| 17 |
+
"revision": null,
|
| 18 |
+
"target_modules": [
|
| 19 |
+
"k_proj",
|
| 20 |
+
"o_proj",
|
| 21 |
+
"v_proj",
|
| 22 |
+
"q_proj"
|
| 23 |
+
],
|
| 24 |
+
"task_type": "CAUSAL_LM"
|
| 25 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c3cb617f4f26f59b2adac4da811c4ddcf2516a14593f17af805749d5ecbf3f92
|
| 3 |
+
size 218138576
|
all_results.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.617767333984375,
|
| 4 |
+
"eval_logits/rejected": -2.721874952316284,
|
| 5 |
+
"eval_logps/chosen": -238.54788208007812,
|
| 6 |
+
"eval_logps/rejected": -388.59033203125,
|
| 7 |
+
"eval_loss": 0.20815864205360413,
|
| 8 |
+
"eval_rewards/accuracies": 0.9413930773735046,
|
| 9 |
+
"eval_rewards/chosen": 1.3856867551803589,
|
| 10 |
+
"eval_rewards/margins": 2.292266845703125,
|
| 11 |
+
"eval_rewards/rejected": -0.9065799117088318,
|
| 12 |
+
"eval_runtime": 2791.9998,
|
| 13 |
+
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 3.403,
|
| 15 |
+
"eval_steps_per_second": 0.426,
|
| 16 |
+
"total_flos": 0.0,
|
| 17 |
+
"train_loss": 0.33413780781687524,
|
| 18 |
+
"train_runtime": 91396.7242,
|
| 19 |
+
"train_samples": 61966,
|
| 20 |
+
"train_samples_per_second": 2.06,
|
| 21 |
+
"train_steps_per_second": 0.016
|
| 22 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_logits/chosen": -2.617767333984375,
|
| 4 |
+
"eval_logits/rejected": -2.721874952316284,
|
| 5 |
+
"eval_logps/chosen": -238.54788208007812,
|
| 6 |
+
"eval_logps/rejected": -388.59033203125,
|
| 7 |
+
"eval_loss": 0.20815864205360413,
|
| 8 |
+
"eval_rewards/accuracies": 0.9413930773735046,
|
| 9 |
+
"eval_rewards/chosen": 1.3856867551803589,
|
| 10 |
+
"eval_rewards/margins": 2.292266845703125,
|
| 11 |
+
"eval_rewards/rejected": -0.9065799117088318,
|
| 12 |
+
"eval_runtime": 2791.9998,
|
| 13 |
+
"eval_samples": 2000,
|
| 14 |
+
"eval_samples_per_second": 3.403,
|
| 15 |
+
"eval_steps_per_second": 0.426
|
| 16 |
+
}
|
runs/Jan05_23-27-01_amaterasu/events.out.tfevents.1704497299.amaterasu.1124388.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c179ca7521c85f957d90d16cd0c6ef86a67e11a79827f4faf5b07c8638edaafc
|
| 3 |
+
size 4424
|
runs/Jan05_23-32-55_amaterasu/events.out.tfevents.1704497659.amaterasu.1124388.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:332afb430f6d83b9d8bc265a65d5b92f9874d1d9335501017307e8445e477b2d
|
| 3 |
+
size 4425
|
runs/Jan05_23-39-40_amaterasu/events.out.tfevents.1704498064.amaterasu.1124388.3
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ec0c4508be3044b5cc6d78e1abe9a6ee56bdfac01ae4072b90a8a834747c0401
|
| 3 |
+
size 4425
|
runs/Jan05_23-47-10_amaterasu/events.out.tfevents.1704498636.amaterasu.1197664.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a84d4e16949b0246722c71e78a55e444e69fb7126a17d3f5ecbc2b11c2b125de
|
| 3 |
+
size 5604
|
runs/Jan05_23-47-10_amaterasu/events.out.tfevents.1704499383.amaterasu.1197664.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:67fc7098474eb2683a998b5e2abe5d75f4b2bb5c2aea866112c592487dda88fa
|
| 3 |
+
size 4397
|
runs/Jan06_00-04-38_amaterasu/events.out.tfevents.1704499547.amaterasu.1204821.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a6b0f195488abb9e5a95a9199b022def032d8bbebe187137dc58b1b2dec4655
|
| 3 |
+
size 4397
|
runs/Jan06_00-08-44_amaterasu/events.out.tfevents.1704499811.amaterasu.1206551.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2660a0abf6e7d06d2e70e04b500b3e4a3140fc44a19cbc6575a943320566ae5c
|
| 3 |
+
size 4424
|
runs/Jan06_00-12-03_amaterasu/events.out.tfevents.1704500027.amaterasu.1207814.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:797411fc6fd5c3684040b9ed4898970d6056bcf5cf9679c11528e7b20b02cb11
|
| 3 |
+
size 99141
|
runs/Jan06_00-12-03_amaterasu/events.out.tfevents.1704594215.amaterasu.1207814.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0520ffa6df0acf10079889746643a4f85aaebd2de7e049fc0b920c5ed6e852e9
|
| 3 |
+
size 828
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "</s>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "<unk>",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "<s>",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "</s>",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
}
|
| 27 |
+
},
|
| 28 |
+
"additional_special_tokens": [],
|
| 29 |
+
"bos_token": "<s>",
|
| 30 |
+
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
| 31 |
+
"clean_up_tokenization_spaces": false,
|
| 32 |
+
"eos_token": "</s>",
|
| 33 |
+
"legacy": true,
|
| 34 |
+
"model_max_length": 2048,
|
| 35 |
+
"pad_token": "</s>",
|
| 36 |
+
"sp_model_kwargs": {},
|
| 37 |
+
"spaces_between_special_tokens": false,
|
| 38 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 39 |
+
"unk_token": "<unk>",
|
| 40 |
+
"use_default_system_prompt": true
|
| 41 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"total_flos": 0.0,
|
| 4 |
+
"train_loss": 0.33413780781687524,
|
| 5 |
+
"train_runtime": 91396.7242,
|
| 6 |
+
"train_samples": 61966,
|
| 7 |
+
"train_samples_per_second": 2.06,
|
| 8 |
+
"train_steps_per_second": 0.016
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2116 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9993414204074695,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 1470,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0,
|
| 13 |
+
"learning_rate": 3.4013605442176867e-09,
|
| 14 |
+
"logits/chosen": -2.8035497665405273,
|
| 15 |
+
"logits/rejected": -2.7962629795074463,
|
| 16 |
+
"logps/chosen": -211.36532592773438,
|
| 17 |
+
"logps/rejected": -294.74530029296875,
|
| 18 |
+
"loss": 0.693,
|
| 19 |
+
"rewards/accuracies": 0.2265625,
|
| 20 |
+
"rewards/chosen": 0.0010320872534066439,
|
| 21 |
+
"rewards/margins": 0.0005493065109476447,
|
| 22 |
+
"rewards/rejected": 0.0004827805096283555,
|
| 23 |
+
"step": 1
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01,
|
| 27 |
+
"learning_rate": 3.4013605442176873e-08,
|
| 28 |
+
"logits/chosen": -2.7791833877563477,
|
| 29 |
+
"logits/rejected": -2.804030418395996,
|
| 30 |
+
"logps/chosen": -240.9124298095703,
|
| 31 |
+
"logps/rejected": -369.5000305175781,
|
| 32 |
+
"loss": 0.6926,
|
| 33 |
+
"rewards/accuracies": 0.4696180522441864,
|
| 34 |
+
"rewards/chosen": 0.00038262151065282524,
|
| 35 |
+
"rewards/margins": 0.0016919042682275176,
|
| 36 |
+
"rewards/rejected": -0.00130928261205554,
|
| 37 |
+
"step": 10
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.01,
|
| 41 |
+
"learning_rate": 6.802721088435375e-08,
|
| 42 |
+
"logits/chosen": -2.7648768424987793,
|
| 43 |
+
"logits/rejected": -2.78273606300354,
|
| 44 |
+
"logps/chosen": -245.15121459960938,
|
| 45 |
+
"logps/rejected": -350.14898681640625,
|
| 46 |
+
"loss": 0.6932,
|
| 47 |
+
"rewards/accuracies": 0.5023437738418579,
|
| 48 |
+
"rewards/chosen": 0.0015446910401806235,
|
| 49 |
+
"rewards/margins": 0.0005673653213307261,
|
| 50 |
+
"rewards/rejected": 0.000977325951680541,
|
| 51 |
+
"step": 20
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.02,
|
| 55 |
+
"learning_rate": 1.0204081632653061e-07,
|
| 56 |
+
"logits/chosen": -2.8178772926330566,
|
| 57 |
+
"logits/rejected": -2.786083221435547,
|
| 58 |
+
"logps/chosen": -240.51516723632812,
|
| 59 |
+
"logps/rejected": -352.41339111328125,
|
| 60 |
+
"loss": 0.6929,
|
| 61 |
+
"rewards/accuracies": 0.508593738079071,
|
| 62 |
+
"rewards/chosen": 0.0015831931959837675,
|
| 63 |
+
"rewards/margins": 0.001221821061335504,
|
| 64 |
+
"rewards/rejected": 0.00036137248389422894,
|
| 65 |
+
"step": 30
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.03,
|
| 69 |
+
"learning_rate": 1.360544217687075e-07,
|
| 70 |
+
"logits/chosen": -2.803492307662964,
|
| 71 |
+
"logits/rejected": -2.7716286182403564,
|
| 72 |
+
"logps/chosen": -235.7887725830078,
|
| 73 |
+
"logps/rejected": -359.8059997558594,
|
| 74 |
+
"loss": 0.6912,
|
| 75 |
+
"rewards/accuracies": 0.5234375,
|
| 76 |
+
"rewards/chosen": 0.0033938586711883545,
|
| 77 |
+
"rewards/margins": 0.004743899218738079,
|
| 78 |
+
"rewards/rejected": -0.001350040198303759,
|
| 79 |
+
"step": 40
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.03,
|
| 83 |
+
"learning_rate": 1.7006802721088434e-07,
|
| 84 |
+
"logits/chosen": -2.8103866577148438,
|
| 85 |
+
"logits/rejected": -2.803828716278076,
|
| 86 |
+
"logps/chosen": -245.4801483154297,
|
| 87 |
+
"logps/rejected": -306.09783935546875,
|
| 88 |
+
"loss": 0.6899,
|
| 89 |
+
"rewards/accuracies": 0.5640624761581421,
|
| 90 |
+
"rewards/chosen": 0.00749587407335639,
|
| 91 |
+
"rewards/margins": 0.007244518492370844,
|
| 92 |
+
"rewards/rejected": 0.00025135590112768114,
|
| 93 |
+
"step": 50
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.04,
|
| 97 |
+
"learning_rate": 2.0408163265306121e-07,
|
| 98 |
+
"logits/chosen": -2.7881524562835693,
|
| 99 |
+
"logits/rejected": -2.808814525604248,
|
| 100 |
+
"logps/chosen": -269.1226501464844,
|
| 101 |
+
"logps/rejected": -339.7763977050781,
|
| 102 |
+
"loss": 0.6874,
|
| 103 |
+
"rewards/accuracies": 0.6015625,
|
| 104 |
+
"rewards/chosen": 0.011595133692026138,
|
| 105 |
+
"rewards/margins": 0.012225830927491188,
|
| 106 |
+
"rewards/rejected": -0.000630697060842067,
|
| 107 |
+
"step": 60
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.05,
|
| 111 |
+
"learning_rate": 2.3809523809523806e-07,
|
| 112 |
+
"logits/chosen": -2.7982544898986816,
|
| 113 |
+
"logits/rejected": -2.765774726867676,
|
| 114 |
+
"logps/chosen": -258.89117431640625,
|
| 115 |
+
"logps/rejected": -372.06451416015625,
|
| 116 |
+
"loss": 0.6864,
|
| 117 |
+
"rewards/accuracies": 0.6148437261581421,
|
| 118 |
+
"rewards/chosen": 0.013396549038589,
|
| 119 |
+
"rewards/margins": 0.014446373097598553,
|
| 120 |
+
"rewards/rejected": -0.0010498259216547012,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.05,
|
| 125 |
+
"learning_rate": 2.72108843537415e-07,
|
| 126 |
+
"logits/chosen": -2.8109402656555176,
|
| 127 |
+
"logits/rejected": -2.7843804359436035,
|
| 128 |
+
"logps/chosen": -244.15817260742188,
|
| 129 |
+
"logps/rejected": -369.6734313964844,
|
| 130 |
+
"loss": 0.6821,
|
| 131 |
+
"rewards/accuracies": 0.6742187738418579,
|
| 132 |
+
"rewards/chosen": 0.02298940345644951,
|
| 133 |
+
"rewards/margins": 0.023194540292024612,
|
| 134 |
+
"rewards/rejected": -0.0002051351184491068,
|
| 135 |
+
"step": 80
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.06,
|
| 139 |
+
"learning_rate": 3.0612244897959183e-07,
|
| 140 |
+
"logits/chosen": -2.8090157508850098,
|
| 141 |
+
"logits/rejected": -2.7707672119140625,
|
| 142 |
+
"logps/chosen": -222.1091766357422,
|
| 143 |
+
"logps/rejected": -365.6192321777344,
|
| 144 |
+
"loss": 0.6766,
|
| 145 |
+
"rewards/accuracies": 0.7242187261581421,
|
| 146 |
+
"rewards/chosen": 0.030939970165491104,
|
| 147 |
+
"rewards/margins": 0.03436826914548874,
|
| 148 |
+
"rewards/rejected": -0.0034283031709492207,
|
| 149 |
+
"step": 90
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.07,
|
| 153 |
+
"learning_rate": 3.401360544217687e-07,
|
| 154 |
+
"logits/chosen": -2.7735049724578857,
|
| 155 |
+
"logits/rejected": -2.7935452461242676,
|
| 156 |
+
"logps/chosen": -251.73049926757812,
|
| 157 |
+
"logps/rejected": -388.00115966796875,
|
| 158 |
+
"loss": 0.6728,
|
| 159 |
+
"rewards/accuracies": 0.735156238079071,
|
| 160 |
+
"rewards/chosen": 0.03729977086186409,
|
| 161 |
+
"rewards/margins": 0.04232599213719368,
|
| 162 |
+
"rewards/rejected": -0.0050262222066521645,
|
| 163 |
+
"step": 100
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.07,
|
| 167 |
+
"learning_rate": 3.741496598639456e-07,
|
| 168 |
+
"logits/chosen": -2.797628164291382,
|
| 169 |
+
"logits/rejected": -2.784834384918213,
|
| 170 |
+
"logps/chosen": -255.72265625,
|
| 171 |
+
"logps/rejected": -349.15985107421875,
|
| 172 |
+
"loss": 0.6651,
|
| 173 |
+
"rewards/accuracies": 0.788281261920929,
|
| 174 |
+
"rewards/chosen": 0.05175922438502312,
|
| 175 |
+
"rewards/margins": 0.05847715586423874,
|
| 176 |
+
"rewards/rejected": -0.00671793520450592,
|
| 177 |
+
"step": 110
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.08,
|
| 181 |
+
"learning_rate": 4.0816326530612243e-07,
|
| 182 |
+
"logits/chosen": -2.7973737716674805,
|
| 183 |
+
"logits/rejected": -2.7825686931610107,
|
| 184 |
+
"logps/chosen": -252.3303985595703,
|
| 185 |
+
"logps/rejected": -348.4207458496094,
|
| 186 |
+
"loss": 0.6604,
|
| 187 |
+
"rewards/accuracies": 0.813281238079071,
|
| 188 |
+
"rewards/chosen": 0.06004839017987251,
|
| 189 |
+
"rewards/margins": 0.06873828917741776,
|
| 190 |
+
"rewards/rejected": -0.008689895272254944,
|
| 191 |
+
"step": 120
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.09,
|
| 195 |
+
"learning_rate": 4.421768707482993e-07,
|
| 196 |
+
"logits/chosen": -2.7856903076171875,
|
| 197 |
+
"logits/rejected": -2.8103625774383545,
|
| 198 |
+
"logps/chosen": -248.4453125,
|
| 199 |
+
"logps/rejected": -316.520263671875,
|
| 200 |
+
"loss": 0.6528,
|
| 201 |
+
"rewards/accuracies": 0.8179687261581421,
|
| 202 |
+
"rewards/chosen": 0.07609430700540543,
|
| 203 |
+
"rewards/margins": 0.08578468859195709,
|
| 204 |
+
"rewards/rejected": -0.00969038438051939,
|
| 205 |
+
"step": 130
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.1,
|
| 209 |
+
"learning_rate": 4.761904761904761e-07,
|
| 210 |
+
"logits/chosen": -2.7964794635772705,
|
| 211 |
+
"logits/rejected": -2.8038413524627686,
|
| 212 |
+
"logps/chosen": -251.0780029296875,
|
| 213 |
+
"logps/rejected": -380.4024353027344,
|
| 214 |
+
"loss": 0.6409,
|
| 215 |
+
"rewards/accuracies": 0.842968761920929,
|
| 216 |
+
"rewards/chosen": 0.10089793056249619,
|
| 217 |
+
"rewards/margins": 0.11140058934688568,
|
| 218 |
+
"rewards/rejected": -0.010502668097615242,
|
| 219 |
+
"step": 140
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.1,
|
| 223 |
+
"learning_rate": 4.988662131519274e-07,
|
| 224 |
+
"logits/chosen": -2.7733452320098877,
|
| 225 |
+
"logits/rejected": -2.799926280975342,
|
| 226 |
+
"logps/chosen": -259.34686279296875,
|
| 227 |
+
"logps/rejected": -335.1527404785156,
|
| 228 |
+
"loss": 0.6297,
|
| 229 |
+
"rewards/accuracies": 0.8539062738418579,
|
| 230 |
+
"rewards/chosen": 0.12008102238178253,
|
| 231 |
+
"rewards/margins": 0.13700444996356964,
|
| 232 |
+
"rewards/rejected": -0.016923416405916214,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.11,
|
| 237 |
+
"learning_rate": 4.950869236583522e-07,
|
| 238 |
+
"logits/chosen": -2.774165153503418,
|
| 239 |
+
"logits/rejected": -2.7881526947021484,
|
| 240 |
+
"logps/chosen": -245.5338134765625,
|
| 241 |
+
"logps/rejected": -338.31597900390625,
|
| 242 |
+
"loss": 0.6201,
|
| 243 |
+
"rewards/accuracies": 0.8687499761581421,
|
| 244 |
+
"rewards/chosen": 0.14051470160484314,
|
| 245 |
+
"rewards/margins": 0.1599283218383789,
|
| 246 |
+
"rewards/rejected": -0.019413620233535767,
|
| 247 |
+
"step": 160
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.12,
|
| 251 |
+
"learning_rate": 4.91307634164777e-07,
|
| 252 |
+
"logits/chosen": -2.811603546142578,
|
| 253 |
+
"logits/rejected": -2.8174936771392822,
|
| 254 |
+
"logps/chosen": -260.7558898925781,
|
| 255 |
+
"logps/rejected": -356.88153076171875,
|
| 256 |
+
"loss": 0.6041,
|
| 257 |
+
"rewards/accuracies": 0.875,
|
| 258 |
+
"rewards/chosen": 0.1749168038368225,
|
| 259 |
+
"rewards/margins": 0.19711166620254517,
|
| 260 |
+
"rewards/rejected": -0.022194867953658104,
|
| 261 |
+
"step": 170
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.12,
|
| 265 |
+
"learning_rate": 4.875283446712018e-07,
|
| 266 |
+
"logits/chosen": -2.7915146350860596,
|
| 267 |
+
"logits/rejected": -2.7889480590820312,
|
| 268 |
+
"logps/chosen": -264.36138916015625,
|
| 269 |
+
"logps/rejected": -353.7435607910156,
|
| 270 |
+
"loss": 0.5926,
|
| 271 |
+
"rewards/accuracies": 0.883593738079071,
|
| 272 |
+
"rewards/chosen": 0.19911792874336243,
|
| 273 |
+
"rewards/margins": 0.22633683681488037,
|
| 274 |
+
"rewards/rejected": -0.02721891924738884,
|
| 275 |
+
"step": 180
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.13,
|
| 279 |
+
"learning_rate": 4.837490551776266e-07,
|
| 280 |
+
"logits/chosen": -2.7990036010742188,
|
| 281 |
+
"logits/rejected": -2.7916808128356934,
|
| 282 |
+
"logps/chosen": -257.4069519042969,
|
| 283 |
+
"logps/rejected": -372.6297302246094,
|
| 284 |
+
"loss": 0.5799,
|
| 285 |
+
"rewards/accuracies": 0.887499988079071,
|
| 286 |
+
"rewards/chosen": 0.22631244361400604,
|
| 287 |
+
"rewards/margins": 0.2581940293312073,
|
| 288 |
+
"rewards/rejected": -0.031881578266620636,
|
| 289 |
+
"step": 190
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.14,
|
| 293 |
+
"learning_rate": 4.799697656840514e-07,
|
| 294 |
+
"logits/chosen": -2.7753312587738037,
|
| 295 |
+
"logits/rejected": -2.7730696201324463,
|
| 296 |
+
"logps/chosen": -259.2568054199219,
|
| 297 |
+
"logps/rejected": -390.26995849609375,
|
| 298 |
+
"loss": 0.564,
|
| 299 |
+
"rewards/accuracies": 0.889843761920929,
|
| 300 |
+
"rewards/chosen": 0.25861743092536926,
|
| 301 |
+
"rewards/margins": 0.30055442452430725,
|
| 302 |
+
"rewards/rejected": -0.0419369637966156,
|
| 303 |
+
"step": 200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.14,
|
| 307 |
+
"learning_rate": 4.761904761904761e-07,
|
| 308 |
+
"logits/chosen": -2.7830989360809326,
|
| 309 |
+
"logits/rejected": -2.7885472774505615,
|
| 310 |
+
"logps/chosen": -229.49685668945312,
|
| 311 |
+
"logps/rejected": -346.35784912109375,
|
| 312 |
+
"loss": 0.5551,
|
| 313 |
+
"rewards/accuracies": 0.91015625,
|
| 314 |
+
"rewards/chosen": 0.28561651706695557,
|
| 315 |
+
"rewards/margins": 0.32180091738700867,
|
| 316 |
+
"rewards/rejected": -0.03618443384766579,
|
| 317 |
+
"step": 210
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.15,
|
| 321 |
+
"learning_rate": 4.7241118669690096e-07,
|
| 322 |
+
"logits/chosen": -2.7914628982543945,
|
| 323 |
+
"logits/rejected": -2.7812819480895996,
|
| 324 |
+
"logps/chosen": -277.1968078613281,
|
| 325 |
+
"logps/rejected": -334.34124755859375,
|
| 326 |
+
"loss": 0.5473,
|
| 327 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 328 |
+
"rewards/chosen": 0.30997538566589355,
|
| 329 |
+
"rewards/margins": 0.3486320972442627,
|
| 330 |
+
"rewards/rejected": -0.038656704127788544,
|
| 331 |
+
"step": 220
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.16,
|
| 335 |
+
"learning_rate": 4.6863189720332574e-07,
|
| 336 |
+
"logits/chosen": -2.7915186882019043,
|
| 337 |
+
"logits/rejected": -2.7635109424591064,
|
| 338 |
+
"logps/chosen": -230.6345672607422,
|
| 339 |
+
"logps/rejected": -366.45855712890625,
|
| 340 |
+
"loss": 0.5283,
|
| 341 |
+
"rewards/accuracies": 0.901562511920929,
|
| 342 |
+
"rewards/chosen": 0.3480406403541565,
|
| 343 |
+
"rewards/margins": 0.3980127274990082,
|
| 344 |
+
"rewards/rejected": -0.04997207969427109,
|
| 345 |
+
"step": 230
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.16,
|
| 349 |
+
"learning_rate": 4.648526077097505e-07,
|
| 350 |
+
"logits/chosen": -2.8176677227020264,
|
| 351 |
+
"logits/rejected": -2.8094589710235596,
|
| 352 |
+
"logps/chosen": -255.73318481445312,
|
| 353 |
+
"logps/rejected": -356.473876953125,
|
| 354 |
+
"loss": 0.5141,
|
| 355 |
+
"rewards/accuracies": 0.905468761920929,
|
| 356 |
+
"rewards/chosen": 0.38035809993743896,
|
| 357 |
+
"rewards/margins": 0.4426742494106293,
|
| 358 |
+
"rewards/rejected": -0.062316179275512695,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.17,
|
| 363 |
+
"learning_rate": 4.6107331821617536e-07,
|
| 364 |
+
"logits/chosen": -2.778831958770752,
|
| 365 |
+
"logits/rejected": -2.7532734870910645,
|
| 366 |
+
"logps/chosen": -260.0787658691406,
|
| 367 |
+
"logps/rejected": -382.69403076171875,
|
| 368 |
+
"loss": 0.5037,
|
| 369 |
+
"rewards/accuracies": 0.9078124761581421,
|
| 370 |
+
"rewards/chosen": 0.4094300866127014,
|
| 371 |
+
"rewards/margins": 0.4735100269317627,
|
| 372 |
+
"rewards/rejected": -0.06407993286848068,
|
| 373 |
+
"step": 250
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.18,
|
| 377 |
+
"learning_rate": 4.5729402872260014e-07,
|
| 378 |
+
"logits/chosen": -2.7875959873199463,
|
| 379 |
+
"logits/rejected": -2.789522647857666,
|
| 380 |
+
"logps/chosen": -245.36215209960938,
|
| 381 |
+
"logps/rejected": -398.8630676269531,
|
| 382 |
+
"loss": 0.4946,
|
| 383 |
+
"rewards/accuracies": 0.897656261920929,
|
| 384 |
+
"rewards/chosen": 0.43164581060409546,
|
| 385 |
+
"rewards/margins": 0.506696879863739,
|
| 386 |
+
"rewards/rejected": -0.07505108416080475,
|
| 387 |
+
"step": 260
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.18,
|
| 391 |
+
"learning_rate": 4.535147392290249e-07,
|
| 392 |
+
"logits/chosen": -2.7784600257873535,
|
| 393 |
+
"logits/rejected": -2.743320941925049,
|
| 394 |
+
"logps/chosen": -240.0518035888672,
|
| 395 |
+
"logps/rejected": -373.5130920410156,
|
| 396 |
+
"loss": 0.4891,
|
| 397 |
+
"rewards/accuracies": 0.89453125,
|
| 398 |
+
"rewards/chosen": 0.45601949095726013,
|
| 399 |
+
"rewards/margins": 0.5297552347183228,
|
| 400 |
+
"rewards/rejected": -0.07373576611280441,
|
| 401 |
+
"step": 270
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.19,
|
| 405 |
+
"learning_rate": 4.497354497354497e-07,
|
| 406 |
+
"logits/chosen": -2.777036190032959,
|
| 407 |
+
"logits/rejected": -2.7678191661834717,
|
| 408 |
+
"logps/chosen": -264.9656677246094,
|
| 409 |
+
"logps/rejected": -373.12042236328125,
|
| 410 |
+
"loss": 0.4766,
|
| 411 |
+
"rewards/accuracies": 0.9156249761581421,
|
| 412 |
+
"rewards/chosen": 0.47401291131973267,
|
| 413 |
+
"rewards/margins": 0.5673891305923462,
|
| 414 |
+
"rewards/rejected": -0.09337621927261353,
|
| 415 |
+
"step": 280
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.2,
|
| 419 |
+
"learning_rate": 4.459561602418745e-07,
|
| 420 |
+
"logits/chosen": -2.7813751697540283,
|
| 421 |
+
"logits/rejected": -2.7827224731445312,
|
| 422 |
+
"logps/chosen": -239.7397918701172,
|
| 423 |
+
"logps/rejected": -392.6272888183594,
|
| 424 |
+
"loss": 0.4603,
|
| 425 |
+
"rewards/accuracies": 0.9117187261581421,
|
| 426 |
+
"rewards/chosen": 0.5112585425376892,
|
| 427 |
+
"rewards/margins": 0.6238077878952026,
|
| 428 |
+
"rewards/rejected": -0.11254926025867462,
|
| 429 |
+
"step": 290
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.2,
|
| 433 |
+
"learning_rate": 4.421768707482993e-07,
|
| 434 |
+
"logits/chosen": -2.784381628036499,
|
| 435 |
+
"logits/rejected": -2.7823455333709717,
|
| 436 |
+
"logps/chosen": -247.23696899414062,
|
| 437 |
+
"logps/rejected": -340.01971435546875,
|
| 438 |
+
"loss": 0.4569,
|
| 439 |
+
"rewards/accuracies": 0.909375011920929,
|
| 440 |
+
"rewards/chosen": 0.5431731939315796,
|
| 441 |
+
"rewards/margins": 0.6343038082122803,
|
| 442 |
+
"rewards/rejected": -0.09113059937953949,
|
| 443 |
+
"step": 300
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.21,
|
| 447 |
+
"learning_rate": 4.383975812547241e-07,
|
| 448 |
+
"logits/chosen": -2.7919013500213623,
|
| 449 |
+
"logits/rejected": -2.7927372455596924,
|
| 450 |
+
"logps/chosen": -244.9982147216797,
|
| 451 |
+
"logps/rejected": -345.5526428222656,
|
| 452 |
+
"loss": 0.4422,
|
| 453 |
+
"rewards/accuracies": 0.922656238079071,
|
| 454 |
+
"rewards/chosen": 0.5760600566864014,
|
| 455 |
+
"rewards/margins": 0.6899352669715881,
|
| 456 |
+
"rewards/rejected": -0.11387525498867035,
|
| 457 |
+
"step": 310
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.22,
|
| 461 |
+
"learning_rate": 4.346182917611489e-07,
|
| 462 |
+
"logits/chosen": -2.786698341369629,
|
| 463 |
+
"logits/rejected": -2.7934978008270264,
|
| 464 |
+
"logps/chosen": -255.37142944335938,
|
| 465 |
+
"logps/rejected": -399.12957763671875,
|
| 466 |
+
"loss": 0.4344,
|
| 467 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 468 |
+
"rewards/chosen": 0.5843140482902527,
|
| 469 |
+
"rewards/margins": 0.7244275808334351,
|
| 470 |
+
"rewards/rejected": -0.1401134431362152,
|
| 471 |
+
"step": 320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.22,
|
| 475 |
+
"learning_rate": 4.308390022675737e-07,
|
| 476 |
+
"logits/chosen": -2.7745113372802734,
|
| 477 |
+
"logits/rejected": -2.7805206775665283,
|
| 478 |
+
"logps/chosen": -252.92514038085938,
|
| 479 |
+
"logps/rejected": -392.51165771484375,
|
| 480 |
+
"loss": 0.4332,
|
| 481 |
+
"rewards/accuracies": 0.9046875238418579,
|
| 482 |
+
"rewards/chosen": 0.5970828533172607,
|
| 483 |
+
"rewards/margins": 0.7323796153068542,
|
| 484 |
+
"rewards/rejected": -0.1352967619895935,
|
| 485 |
+
"step": 330
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.23,
|
| 489 |
+
"learning_rate": 4.270597127739985e-07,
|
| 490 |
+
"logits/chosen": -2.783926486968994,
|
| 491 |
+
"logits/rejected": -2.7875866889953613,
|
| 492 |
+
"logps/chosen": -250.8353729248047,
|
| 493 |
+
"logps/rejected": -335.03265380859375,
|
| 494 |
+
"loss": 0.4175,
|
| 495 |
+
"rewards/accuracies": 0.9164062738418579,
|
| 496 |
+
"rewards/chosen": 0.6492675542831421,
|
| 497 |
+
"rewards/margins": 0.7875067591667175,
|
| 498 |
+
"rewards/rejected": -0.13823917508125305,
|
| 499 |
+
"step": 340
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.24,
|
| 503 |
+
"learning_rate": 4.2328042328042324e-07,
|
| 504 |
+
"logits/chosen": -2.7828190326690674,
|
| 505 |
+
"logits/rejected": -2.772052764892578,
|
| 506 |
+
"logps/chosen": -236.33706665039062,
|
| 507 |
+
"logps/rejected": -370.28399658203125,
|
| 508 |
+
"loss": 0.4152,
|
| 509 |
+
"rewards/accuracies": 0.9242187738418579,
|
| 510 |
+
"rewards/chosen": 0.662378191947937,
|
| 511 |
+
"rewards/margins": 0.7986767888069153,
|
| 512 |
+
"rewards/rejected": -0.13629861176013947,
|
| 513 |
+
"step": 350
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.24,
|
| 517 |
+
"learning_rate": 4.19501133786848e-07,
|
| 518 |
+
"logits/chosen": -2.780648946762085,
|
| 519 |
+
"logits/rejected": -2.771820545196533,
|
| 520 |
+
"logps/chosen": -228.22445678710938,
|
| 521 |
+
"logps/rejected": -390.63751220703125,
|
| 522 |
+
"loss": 0.4051,
|
| 523 |
+
"rewards/accuracies": 0.9140625,
|
| 524 |
+
"rewards/chosen": 0.6962443590164185,
|
| 525 |
+
"rewards/margins": 0.8446155786514282,
|
| 526 |
+
"rewards/rejected": -0.1483711302280426,
|
| 527 |
+
"step": 360
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.25,
|
| 531 |
+
"learning_rate": 4.1572184429327286e-07,
|
| 532 |
+
"logits/chosen": -2.8088645935058594,
|
| 533 |
+
"logits/rejected": -2.7826154232025146,
|
| 534 |
+
"logps/chosen": -255.2318572998047,
|
| 535 |
+
"logps/rejected": -344.69183349609375,
|
| 536 |
+
"loss": 0.3908,
|
| 537 |
+
"rewards/accuracies": 0.9203125238418579,
|
| 538 |
+
"rewards/chosen": 0.7306076288223267,
|
| 539 |
+
"rewards/margins": 0.900040328502655,
|
| 540 |
+
"rewards/rejected": -0.16943258047103882,
|
| 541 |
+
"step": 370
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.26,
|
| 545 |
+
"learning_rate": 4.1194255479969764e-07,
|
| 546 |
+
"logits/chosen": -2.7837393283843994,
|
| 547 |
+
"logits/rejected": -2.754739284515381,
|
| 548 |
+
"logps/chosen": -252.39779663085938,
|
| 549 |
+
"logps/rejected": -347.7734069824219,
|
| 550 |
+
"loss": 0.4019,
|
| 551 |
+
"rewards/accuracies": 0.907031238079071,
|
| 552 |
+
"rewards/chosen": 0.7146260738372803,
|
| 553 |
+
"rewards/margins": 0.8642898797988892,
|
| 554 |
+
"rewards/rejected": -0.14966385066509247,
|
| 555 |
+
"step": 380
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.27,
|
| 559 |
+
"learning_rate": 4.0816326530612243e-07,
|
| 560 |
+
"logits/chosen": -2.793994426727295,
|
| 561 |
+
"logits/rejected": -2.789456605911255,
|
| 562 |
+
"logps/chosen": -250.083984375,
|
| 563 |
+
"logps/rejected": -345.2536315917969,
|
| 564 |
+
"loss": 0.3843,
|
| 565 |
+
"rewards/accuracies": 0.9203125238418579,
|
| 566 |
+
"rewards/chosen": 0.760775089263916,
|
| 567 |
+
"rewards/margins": 0.9284068942070007,
|
| 568 |
+
"rewards/rejected": -0.1676318198442459,
|
| 569 |
+
"step": 390
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.27,
|
| 573 |
+
"learning_rate": 4.0438397581254726e-07,
|
| 574 |
+
"logits/chosen": -2.7863235473632812,
|
| 575 |
+
"logits/rejected": -2.7660741806030273,
|
| 576 |
+
"logps/chosen": -243.2860565185547,
|
| 577 |
+
"logps/rejected": -375.15283203125,
|
| 578 |
+
"loss": 0.3736,
|
| 579 |
+
"rewards/accuracies": 0.9195312261581421,
|
| 580 |
+
"rewards/chosen": 0.7728086113929749,
|
| 581 |
+
"rewards/margins": 0.9798704385757446,
|
| 582 |
+
"rewards/rejected": -0.20706184208393097,
|
| 583 |
+
"step": 400
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.28,
|
| 587 |
+
"learning_rate": 4.0060468631897205e-07,
|
| 588 |
+
"logits/chosen": -2.7740797996520996,
|
| 589 |
+
"logits/rejected": -2.787078857421875,
|
| 590 |
+
"logps/chosen": -231.3814239501953,
|
| 591 |
+
"logps/rejected": -373.4275817871094,
|
| 592 |
+
"loss": 0.3779,
|
| 593 |
+
"rewards/accuracies": 0.9140625,
|
| 594 |
+
"rewards/chosen": 0.786165177822113,
|
| 595 |
+
"rewards/margins": 0.9645744562149048,
|
| 596 |
+
"rewards/rejected": -0.1784091293811798,
|
| 597 |
+
"step": 410
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.29,
|
| 601 |
+
"learning_rate": 3.968253968253968e-07,
|
| 602 |
+
"logits/chosen": -2.7854466438293457,
|
| 603 |
+
"logits/rejected": -2.782599449157715,
|
| 604 |
+
"logps/chosen": -234.27853393554688,
|
| 605 |
+
"logps/rejected": -341.40106201171875,
|
| 606 |
+
"loss": 0.3758,
|
| 607 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 608 |
+
"rewards/chosen": 0.8017100095748901,
|
| 609 |
+
"rewards/margins": 0.9820283651351929,
|
| 610 |
+
"rewards/rejected": -0.1803184449672699,
|
| 611 |
+
"step": 420
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.29,
|
| 615 |
+
"learning_rate": 3.930461073318216e-07,
|
| 616 |
+
"logits/chosen": -2.7634427547454834,
|
| 617 |
+
"logits/rejected": -2.7768495082855225,
|
| 618 |
+
"logps/chosen": -230.73318481445312,
|
| 619 |
+
"logps/rejected": -427.71917724609375,
|
| 620 |
+
"loss": 0.3665,
|
| 621 |
+
"rewards/accuracies": 0.91796875,
|
| 622 |
+
"rewards/chosen": 0.8091424703598022,
|
| 623 |
+
"rewards/margins": 1.025179386138916,
|
| 624 |
+
"rewards/rejected": -0.2160368263721466,
|
| 625 |
+
"step": 430
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.3,
|
| 629 |
+
"learning_rate": 3.892668178382464e-07,
|
| 630 |
+
"logits/chosen": -2.774629592895508,
|
| 631 |
+
"logits/rejected": -2.7814247608184814,
|
| 632 |
+
"logps/chosen": -253.4683074951172,
|
| 633 |
+
"logps/rejected": -386.40216064453125,
|
| 634 |
+
"loss": 0.3495,
|
| 635 |
+
"rewards/accuracies": 0.925000011920929,
|
| 636 |
+
"rewards/chosen": 0.8851088285446167,
|
| 637 |
+
"rewards/margins": 1.123652696609497,
|
| 638 |
+
"rewards/rejected": -0.23854386806488037,
|
| 639 |
+
"step": 440
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.31,
|
| 643 |
+
"learning_rate": 3.854875283446712e-07,
|
| 644 |
+
"logits/chosen": -2.766551971435547,
|
| 645 |
+
"logits/rejected": -2.7709641456604004,
|
| 646 |
+
"logps/chosen": -271.8524475097656,
|
| 647 |
+
"logps/rejected": -379.4809265136719,
|
| 648 |
+
"loss": 0.3575,
|
| 649 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 650 |
+
"rewards/chosen": 0.8432048559188843,
|
| 651 |
+
"rewards/margins": 1.0976295471191406,
|
| 652 |
+
"rewards/rejected": -0.254424512386322,
|
| 653 |
+
"step": 450
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.31,
|
| 657 |
+
"learning_rate": 3.8170823885109596e-07,
|
| 658 |
+
"logits/chosen": -2.8009865283966064,
|
| 659 |
+
"logits/rejected": -2.7705283164978027,
|
| 660 |
+
"logps/chosen": -241.07632446289062,
|
| 661 |
+
"logps/rejected": -366.87127685546875,
|
| 662 |
+
"loss": 0.3459,
|
| 663 |
+
"rewards/accuracies": 0.9281250238418579,
|
| 664 |
+
"rewards/chosen": 0.9000816345214844,
|
| 665 |
+
"rewards/margins": 1.1292930841445923,
|
| 666 |
+
"rewards/rejected": -0.22921133041381836,
|
| 667 |
+
"step": 460
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.32,
|
| 671 |
+
"learning_rate": 3.779289493575208e-07,
|
| 672 |
+
"logits/chosen": -2.7855477333068848,
|
| 673 |
+
"logits/rejected": -2.771469831466675,
|
| 674 |
+
"logps/chosen": -248.2216033935547,
|
| 675 |
+
"logps/rejected": -379.58709716796875,
|
| 676 |
+
"loss": 0.3488,
|
| 677 |
+
"rewards/accuracies": 0.913281261920929,
|
| 678 |
+
"rewards/chosen": 0.8979974985122681,
|
| 679 |
+
"rewards/margins": 1.1383633613586426,
|
| 680 |
+
"rewards/rejected": -0.2403658926486969,
|
| 681 |
+
"step": 470
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.33,
|
| 685 |
+
"learning_rate": 3.741496598639456e-07,
|
| 686 |
+
"logits/chosen": -2.783979892730713,
|
| 687 |
+
"logits/rejected": -2.787400722503662,
|
| 688 |
+
"logps/chosen": -234.78939819335938,
|
| 689 |
+
"logps/rejected": -391.0784912109375,
|
| 690 |
+
"loss": 0.3396,
|
| 691 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 692 |
+
"rewards/chosen": 0.8895782232284546,
|
| 693 |
+
"rewards/margins": 1.1713939905166626,
|
| 694 |
+
"rewards/rejected": -0.281815767288208,
|
| 695 |
+
"step": 480
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.33,
|
| 699 |
+
"learning_rate": 3.703703703703703e-07,
|
| 700 |
+
"logits/chosen": -2.7796401977539062,
|
| 701 |
+
"logits/rejected": -2.78939151763916,
|
| 702 |
+
"logps/chosen": -255.79556274414062,
|
| 703 |
+
"logps/rejected": -376.7617492675781,
|
| 704 |
+
"loss": 0.3496,
|
| 705 |
+
"rewards/accuracies": 0.91796875,
|
| 706 |
+
"rewards/chosen": 0.8880151510238647,
|
| 707 |
+
"rewards/margins": 1.1511998176574707,
|
| 708 |
+
"rewards/rejected": -0.26318463683128357,
|
| 709 |
+
"step": 490
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.34,
|
| 713 |
+
"learning_rate": 3.6659108087679515e-07,
|
| 714 |
+
"logits/chosen": -2.784447193145752,
|
| 715 |
+
"logits/rejected": -2.7811279296875,
|
| 716 |
+
"logps/chosen": -240.26943969726562,
|
| 717 |
+
"logps/rejected": -373.43585205078125,
|
| 718 |
+
"loss": 0.3317,
|
| 719 |
+
"rewards/accuracies": 0.9242187738418579,
|
| 720 |
+
"rewards/chosen": 0.9207477569580078,
|
| 721 |
+
"rewards/margins": 1.2141565084457397,
|
| 722 |
+
"rewards/rejected": -0.2934088110923767,
|
| 723 |
+
"step": 500
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.35,
|
| 727 |
+
"learning_rate": 3.6281179138321993e-07,
|
| 728 |
+
"logits/chosen": -2.7936480045318604,
|
| 729 |
+
"logits/rejected": -2.7741034030914307,
|
| 730 |
+
"logps/chosen": -253.25625610351562,
|
| 731 |
+
"logps/rejected": -388.1740417480469,
|
| 732 |
+
"loss": 0.3307,
|
| 733 |
+
"rewards/accuracies": 0.925000011920929,
|
| 734 |
+
"rewards/chosen": 0.9261225461959839,
|
| 735 |
+
"rewards/margins": 1.2367761135101318,
|
| 736 |
+
"rewards/rejected": -0.310653418302536,
|
| 737 |
+
"step": 510
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.35,
|
| 741 |
+
"learning_rate": 3.590325018896447e-07,
|
| 742 |
+
"logits/chosen": -2.764971971511841,
|
| 743 |
+
"logits/rejected": -2.779900074005127,
|
| 744 |
+
"logps/chosen": -277.50433349609375,
|
| 745 |
+
"logps/rejected": -390.9405822753906,
|
| 746 |
+
"loss": 0.3301,
|
| 747 |
+
"rewards/accuracies": 0.9203125238418579,
|
| 748 |
+
"rewards/chosen": 0.9403823614120483,
|
| 749 |
+
"rewards/margins": 1.250135898590088,
|
| 750 |
+
"rewards/rejected": -0.3097533881664276,
|
| 751 |
+
"step": 520
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.36,
|
| 755 |
+
"learning_rate": 3.5525321239606955e-07,
|
| 756 |
+
"logits/chosen": -2.7859063148498535,
|
| 757 |
+
"logits/rejected": -2.7852673530578613,
|
| 758 |
+
"logps/chosen": -240.83847045898438,
|
| 759 |
+
"logps/rejected": -329.5592346191406,
|
| 760 |
+
"loss": 0.3185,
|
| 761 |
+
"rewards/accuracies": 0.9296875,
|
| 762 |
+
"rewards/chosen": 0.998257040977478,
|
| 763 |
+
"rewards/margins": 1.3062750101089478,
|
| 764 |
+
"rewards/rejected": -0.30801790952682495,
|
| 765 |
+
"step": 530
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.37,
|
| 769 |
+
"learning_rate": 3.5147392290249433e-07,
|
| 770 |
+
"logits/chosen": -2.7856059074401855,
|
| 771 |
+
"logits/rejected": -2.7904558181762695,
|
| 772 |
+
"logps/chosen": -256.13116455078125,
|
| 773 |
+
"logps/rejected": -359.0440673828125,
|
| 774 |
+
"loss": 0.3201,
|
| 775 |
+
"rewards/accuracies": 0.917187511920929,
|
| 776 |
+
"rewards/chosen": 0.9812418222427368,
|
| 777 |
+
"rewards/margins": 1.2980186939239502,
|
| 778 |
+
"rewards/rejected": -0.316776841878891,
|
| 779 |
+
"step": 540
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.37,
|
| 783 |
+
"learning_rate": 3.4769463340891906e-07,
|
| 784 |
+
"logits/chosen": -2.7746291160583496,
|
| 785 |
+
"logits/rejected": -2.8083655834198,
|
| 786 |
+
"logps/chosen": -243.3596649169922,
|
| 787 |
+
"logps/rejected": -381.6620788574219,
|
| 788 |
+
"loss": 0.321,
|
| 789 |
+
"rewards/accuracies": 0.921093761920929,
|
| 790 |
+
"rewards/chosen": 0.9785689115524292,
|
| 791 |
+
"rewards/margins": 1.3102028369903564,
|
| 792 |
+
"rewards/rejected": -0.33163395524024963,
|
| 793 |
+
"step": 550
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.38,
|
| 797 |
+
"learning_rate": 3.439153439153439e-07,
|
| 798 |
+
"logits/chosen": -2.788200616836548,
|
| 799 |
+
"logits/rejected": -2.806088924407959,
|
| 800 |
+
"logps/chosen": -243.46371459960938,
|
| 801 |
+
"logps/rejected": -353.0728454589844,
|
| 802 |
+
"loss": 0.3037,
|
| 803 |
+
"rewards/accuracies": 0.9281250238418579,
|
| 804 |
+
"rewards/chosen": 1.0423057079315186,
|
| 805 |
+
"rewards/margins": 1.40134596824646,
|
| 806 |
+
"rewards/rejected": -0.35903996229171753,
|
| 807 |
+
"step": 560
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.39,
|
| 811 |
+
"learning_rate": 3.401360544217687e-07,
|
| 812 |
+
"logits/chosen": -2.8205642700195312,
|
| 813 |
+
"logits/rejected": -2.75651216506958,
|
| 814 |
+
"logps/chosen": -225.49380493164062,
|
| 815 |
+
"logps/rejected": -383.3102111816406,
|
| 816 |
+
"loss": 0.2961,
|
| 817 |
+
"rewards/accuracies": 0.934374988079071,
|
| 818 |
+
"rewards/chosen": 1.0518951416015625,
|
| 819 |
+
"rewards/margins": 1.416092872619629,
|
| 820 |
+
"rewards/rejected": -0.3641977310180664,
|
| 821 |
+
"step": 570
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.39,
|
| 825 |
+
"learning_rate": 3.3635676492819346e-07,
|
| 826 |
+
"logits/chosen": -2.778111696243286,
|
| 827 |
+
"logits/rejected": -2.8062729835510254,
|
| 828 |
+
"logps/chosen": -241.8183135986328,
|
| 829 |
+
"logps/rejected": -360.12677001953125,
|
| 830 |
+
"loss": 0.3026,
|
| 831 |
+
"rewards/accuracies": 0.928906261920929,
|
| 832 |
+
"rewards/chosen": 1.0336360931396484,
|
| 833 |
+
"rewards/margins": 1.3975627422332764,
|
| 834 |
+
"rewards/rejected": -0.3639264702796936,
|
| 835 |
+
"step": 580
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.4,
|
| 839 |
+
"learning_rate": 3.325774754346183e-07,
|
| 840 |
+
"logits/chosen": -2.7760305404663086,
|
| 841 |
+
"logits/rejected": -2.7639145851135254,
|
| 842 |
+
"logps/chosen": -263.2132568359375,
|
| 843 |
+
"logps/rejected": -326.3753356933594,
|
| 844 |
+
"loss": 0.3079,
|
| 845 |
+
"rewards/accuracies": 0.917187511920929,
|
| 846 |
+
"rewards/chosen": 1.0236365795135498,
|
| 847 |
+
"rewards/margins": 1.3882102966308594,
|
| 848 |
+
"rewards/rejected": -0.36457380652427673,
|
| 849 |
+
"step": 590
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.41,
|
| 853 |
+
"learning_rate": 3.287981859410431e-07,
|
| 854 |
+
"logits/chosen": -2.8040480613708496,
|
| 855 |
+
"logits/rejected": -2.781839370727539,
|
| 856 |
+
"logps/chosen": -232.688720703125,
|
| 857 |
+
"logps/rejected": -341.75372314453125,
|
| 858 |
+
"loss": 0.294,
|
| 859 |
+
"rewards/accuracies": 0.92578125,
|
| 860 |
+
"rewards/chosen": 1.0763448476791382,
|
| 861 |
+
"rewards/margins": 1.459729790687561,
|
| 862 |
+
"rewards/rejected": -0.38338491320610046,
|
| 863 |
+
"step": 600
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.41,
|
| 867 |
+
"learning_rate": 3.2501889644746787e-07,
|
| 868 |
+
"logits/chosen": -2.797874927520752,
|
| 869 |
+
"logits/rejected": -2.748481512069702,
|
| 870 |
+
"logps/chosen": -232.8326873779297,
|
| 871 |
+
"logps/rejected": -369.7907409667969,
|
| 872 |
+
"loss": 0.2837,
|
| 873 |
+
"rewards/accuracies": 0.938281238079071,
|
| 874 |
+
"rewards/chosen": 1.101806640625,
|
| 875 |
+
"rewards/margins": 1.498957633972168,
|
| 876 |
+
"rewards/rejected": -0.39715105295181274,
|
| 877 |
+
"step": 610
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.42,
|
| 881 |
+
"learning_rate": 3.2123960695389265e-07,
|
| 882 |
+
"logits/chosen": -2.780925989151001,
|
| 883 |
+
"logits/rejected": -2.735792636871338,
|
| 884 |
+
"logps/chosen": -222.20596313476562,
|
| 885 |
+
"logps/rejected": -380.5815124511719,
|
| 886 |
+
"loss": 0.2935,
|
| 887 |
+
"rewards/accuracies": 0.921093761920929,
|
| 888 |
+
"rewards/chosen": 1.0859084129333496,
|
| 889 |
+
"rewards/margins": 1.4819860458374023,
|
| 890 |
+
"rewards/rejected": -0.39607763290405273,
|
| 891 |
+
"step": 620
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.43,
|
| 895 |
+
"learning_rate": 3.1746031746031743e-07,
|
| 896 |
+
"logits/chosen": -2.7768056392669678,
|
| 897 |
+
"logits/rejected": -2.764166831970215,
|
| 898 |
+
"logps/chosen": -236.9914093017578,
|
| 899 |
+
"logps/rejected": -345.6325378417969,
|
| 900 |
+
"loss": 0.2895,
|
| 901 |
+
"rewards/accuracies": 0.932812511920929,
|
| 902 |
+
"rewards/chosen": 1.1014459133148193,
|
| 903 |
+
"rewards/margins": 1.5069670677185059,
|
| 904 |
+
"rewards/rejected": -0.40552106499671936,
|
| 905 |
+
"step": 630
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.44,
|
| 909 |
+
"learning_rate": 3.136810279667422e-07,
|
| 910 |
+
"logits/chosen": -2.7987258434295654,
|
| 911 |
+
"logits/rejected": -2.8054118156433105,
|
| 912 |
+
"logps/chosen": -235.97109985351562,
|
| 913 |
+
"logps/rejected": -330.56439208984375,
|
| 914 |
+
"loss": 0.2775,
|
| 915 |
+
"rewards/accuracies": 0.93359375,
|
| 916 |
+
"rewards/chosen": 1.1580729484558105,
|
| 917 |
+
"rewards/margins": 1.5699806213378906,
|
| 918 |
+
"rewards/rejected": -0.4119076728820801,
|
| 919 |
+
"step": 640
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.44,
|
| 923 |
+
"learning_rate": 3.0990173847316705e-07,
|
| 924 |
+
"logits/chosen": -2.7858521938323975,
|
| 925 |
+
"logits/rejected": -2.779346466064453,
|
| 926 |
+
"logps/chosen": -257.5158386230469,
|
| 927 |
+
"logps/rejected": -322.25103759765625,
|
| 928 |
+
"loss": 0.287,
|
| 929 |
+
"rewards/accuracies": 0.9195312261581421,
|
| 930 |
+
"rewards/chosen": 1.1325995922088623,
|
| 931 |
+
"rewards/margins": 1.5360453128814697,
|
| 932 |
+
"rewards/rejected": -0.40344563126564026,
|
| 933 |
+
"step": 650
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.45,
|
| 937 |
+
"learning_rate": 3.0612244897959183e-07,
|
| 938 |
+
"logits/chosen": -2.7976508140563965,
|
| 939 |
+
"logits/rejected": -2.8010151386260986,
|
| 940 |
+
"logps/chosen": -219.1446533203125,
|
| 941 |
+
"logps/rejected": -315.2838439941406,
|
| 942 |
+
"loss": 0.2703,
|
| 943 |
+
"rewards/accuracies": 0.9453125,
|
| 944 |
+
"rewards/chosen": 1.1511547565460205,
|
| 945 |
+
"rewards/margins": 1.5933144092559814,
|
| 946 |
+
"rewards/rejected": -0.44215965270996094,
|
| 947 |
+
"step": 660
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.46,
|
| 951 |
+
"learning_rate": 3.023431594860166e-07,
|
| 952 |
+
"logits/chosen": -2.767582416534424,
|
| 953 |
+
"logits/rejected": -2.8024327754974365,
|
| 954 |
+
"logps/chosen": -237.21578979492188,
|
| 955 |
+
"logps/rejected": -314.68377685546875,
|
| 956 |
+
"loss": 0.2637,
|
| 957 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 958 |
+
"rewards/chosen": 1.1508355140686035,
|
| 959 |
+
"rewards/margins": 1.6350256204605103,
|
| 960 |
+
"rewards/rejected": -0.48419007658958435,
|
| 961 |
+
"step": 670
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.46,
|
| 965 |
+
"learning_rate": 2.9856386999244145e-07,
|
| 966 |
+
"logits/chosen": -2.7926082611083984,
|
| 967 |
+
"logits/rejected": -2.780251979827881,
|
| 968 |
+
"logps/chosen": -244.810302734375,
|
| 969 |
+
"logps/rejected": -347.9936828613281,
|
| 970 |
+
"loss": 0.2784,
|
| 971 |
+
"rewards/accuracies": 0.930468738079071,
|
| 972 |
+
"rewards/chosen": 1.1081712245941162,
|
| 973 |
+
"rewards/margins": 1.5819367170333862,
|
| 974 |
+
"rewards/rejected": -0.47376567125320435,
|
| 975 |
+
"step": 680
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.47,
|
| 979 |
+
"learning_rate": 2.947845804988662e-07,
|
| 980 |
+
"logits/chosen": -2.771953821182251,
|
| 981 |
+
"logits/rejected": -2.768907070159912,
|
| 982 |
+
"logps/chosen": -248.50332641601562,
|
| 983 |
+
"logps/rejected": -360.6126403808594,
|
| 984 |
+
"loss": 0.2897,
|
| 985 |
+
"rewards/accuracies": 0.921875,
|
| 986 |
+
"rewards/chosen": 1.0936378240585327,
|
| 987 |
+
"rewards/margins": 1.5781736373901367,
|
| 988 |
+
"rewards/rejected": -0.4845358729362488,
|
| 989 |
+
"step": 690
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.48,
|
| 993 |
+
"learning_rate": 2.9100529100529097e-07,
|
| 994 |
+
"logits/chosen": -2.7748546600341797,
|
| 995 |
+
"logits/rejected": -2.7857470512390137,
|
| 996 |
+
"logps/chosen": -227.1557159423828,
|
| 997 |
+
"logps/rejected": -390.3030700683594,
|
| 998 |
+
"loss": 0.2597,
|
| 999 |
+
"rewards/accuracies": 0.9320312738418579,
|
| 1000 |
+
"rewards/chosen": 1.1781264543533325,
|
| 1001 |
+
"rewards/margins": 1.7109047174453735,
|
| 1002 |
+
"rewards/rejected": -0.5327781438827515,
|
| 1003 |
+
"step": 700
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.48,
|
| 1007 |
+
"learning_rate": 2.872260015117158e-07,
|
| 1008 |
+
"logits/chosen": -2.77628231048584,
|
| 1009 |
+
"logits/rejected": -2.7869679927825928,
|
| 1010 |
+
"logps/chosen": -245.57839965820312,
|
| 1011 |
+
"logps/rejected": -326.86212158203125,
|
| 1012 |
+
"loss": 0.2613,
|
| 1013 |
+
"rewards/accuracies": 0.934374988079071,
|
| 1014 |
+
"rewards/chosen": 1.1419804096221924,
|
| 1015 |
+
"rewards/margins": 1.6727325916290283,
|
| 1016 |
+
"rewards/rejected": -0.5307522416114807,
|
| 1017 |
+
"step": 710
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.49,
|
| 1021 |
+
"learning_rate": 2.834467120181406e-07,
|
| 1022 |
+
"logits/chosen": -2.7608537673950195,
|
| 1023 |
+
"logits/rejected": -2.7646660804748535,
|
| 1024 |
+
"logps/chosen": -241.5836944580078,
|
| 1025 |
+
"logps/rejected": -379.62860107421875,
|
| 1026 |
+
"loss": 0.2738,
|
| 1027 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1028 |
+
"rewards/chosen": 1.1373337507247925,
|
| 1029 |
+
"rewards/margins": 1.6593284606933594,
|
| 1030 |
+
"rewards/rejected": -0.5219947099685669,
|
| 1031 |
+
"step": 720
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.5,
|
| 1035 |
+
"learning_rate": 2.7966742252456537e-07,
|
| 1036 |
+
"logits/chosen": -2.777465343475342,
|
| 1037 |
+
"logits/rejected": -2.801975965499878,
|
| 1038 |
+
"logps/chosen": -227.2059326171875,
|
| 1039 |
+
"logps/rejected": -369.7891540527344,
|
| 1040 |
+
"loss": 0.2554,
|
| 1041 |
+
"rewards/accuracies": 0.934374988079071,
|
| 1042 |
+
"rewards/chosen": 1.209241271018982,
|
| 1043 |
+
"rewards/margins": 1.7282158136367798,
|
| 1044 |
+
"rewards/rejected": -0.5189744234085083,
|
| 1045 |
+
"step": 730
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.5,
|
| 1049 |
+
"learning_rate": 2.758881330309902e-07,
|
| 1050 |
+
"logits/chosen": -2.7639384269714355,
|
| 1051 |
+
"logits/rejected": -2.7558932304382324,
|
| 1052 |
+
"logps/chosen": -255.972412109375,
|
| 1053 |
+
"logps/rejected": -410.17431640625,
|
| 1054 |
+
"loss": 0.2766,
|
| 1055 |
+
"rewards/accuracies": 0.9156249761581421,
|
| 1056 |
+
"rewards/chosen": 1.1485779285430908,
|
| 1057 |
+
"rewards/margins": 1.672486662864685,
|
| 1058 |
+
"rewards/rejected": -0.5239086151123047,
|
| 1059 |
+
"step": 740
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.51,
|
| 1063 |
+
"learning_rate": 2.72108843537415e-07,
|
| 1064 |
+
"logits/chosen": -2.7429962158203125,
|
| 1065 |
+
"logits/rejected": -2.7603325843811035,
|
| 1066 |
+
"logps/chosen": -248.05697631835938,
|
| 1067 |
+
"logps/rejected": -382.65863037109375,
|
| 1068 |
+
"loss": 0.2692,
|
| 1069 |
+
"rewards/accuracies": 0.934374988079071,
|
| 1070 |
+
"rewards/chosen": 1.1628259420394897,
|
| 1071 |
+
"rewards/margins": 1.7002170085906982,
|
| 1072 |
+
"rewards/rejected": -0.5373910665512085,
|
| 1073 |
+
"step": 750
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.52,
|
| 1077 |
+
"learning_rate": 2.683295540438397e-07,
|
| 1078 |
+
"logits/chosen": -2.7732365131378174,
|
| 1079 |
+
"logits/rejected": -2.7899222373962402,
|
| 1080 |
+
"logps/chosen": -230.82577514648438,
|
| 1081 |
+
"logps/rejected": -356.39349365234375,
|
| 1082 |
+
"loss": 0.262,
|
| 1083 |
+
"rewards/accuracies": 0.93359375,
|
| 1084 |
+
"rewards/chosen": 1.1516262292861938,
|
| 1085 |
+
"rewards/margins": 1.7132419347763062,
|
| 1086 |
+
"rewards/rejected": -0.5616158843040466,
|
| 1087 |
+
"step": 760
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.52,
|
| 1091 |
+
"learning_rate": 2.645502645502645e-07,
|
| 1092 |
+
"logits/chosen": -2.764669895172119,
|
| 1093 |
+
"logits/rejected": -2.7641212940216064,
|
| 1094 |
+
"logps/chosen": -246.3456573486328,
|
| 1095 |
+
"logps/rejected": -370.99896240234375,
|
| 1096 |
+
"loss": 0.2701,
|
| 1097 |
+
"rewards/accuracies": 0.9164062738418579,
|
| 1098 |
+
"rewards/chosen": 1.191197395324707,
|
| 1099 |
+
"rewards/margins": 1.7232650518417358,
|
| 1100 |
+
"rewards/rejected": -0.5320678949356079,
|
| 1101 |
+
"step": 770
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.53,
|
| 1105 |
+
"learning_rate": 2.6077097505668934e-07,
|
| 1106 |
+
"logits/chosen": -2.7817633152008057,
|
| 1107 |
+
"logits/rejected": -2.7922616004943848,
|
| 1108 |
+
"logps/chosen": -256.2757873535156,
|
| 1109 |
+
"logps/rejected": -356.1881408691406,
|
| 1110 |
+
"loss": 0.2571,
|
| 1111 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 1112 |
+
"rewards/chosen": 1.2059863805770874,
|
| 1113 |
+
"rewards/margins": 1.7701711654663086,
|
| 1114 |
+
"rewards/rejected": -0.5641847848892212,
|
| 1115 |
+
"step": 780
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.54,
|
| 1119 |
+
"learning_rate": 2.569916855631141e-07,
|
| 1120 |
+
"logits/chosen": -2.7949161529541016,
|
| 1121 |
+
"logits/rejected": -2.800379514694214,
|
| 1122 |
+
"logps/chosen": -219.2698516845703,
|
| 1123 |
+
"logps/rejected": -384.794189453125,
|
| 1124 |
+
"loss": 0.2512,
|
| 1125 |
+
"rewards/accuracies": 0.9296875,
|
| 1126 |
+
"rewards/chosen": 1.2469325065612793,
|
| 1127 |
+
"rewards/margins": 1.8279892206192017,
|
| 1128 |
+
"rewards/rejected": -0.5810565948486328,
|
| 1129 |
+
"step": 790
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.54,
|
| 1133 |
+
"learning_rate": 2.532123960695389e-07,
|
| 1134 |
+
"logits/chosen": -2.7864224910736084,
|
| 1135 |
+
"logits/rejected": -2.8051304817199707,
|
| 1136 |
+
"logps/chosen": -243.42105102539062,
|
| 1137 |
+
"logps/rejected": -376.7647399902344,
|
| 1138 |
+
"loss": 0.2455,
|
| 1139 |
+
"rewards/accuracies": 0.9398437738418579,
|
| 1140 |
+
"rewards/chosen": 1.25649094581604,
|
| 1141 |
+
"rewards/margins": 1.8448721170425415,
|
| 1142 |
+
"rewards/rejected": -0.5883811712265015,
|
| 1143 |
+
"step": 800
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.55,
|
| 1147 |
+
"learning_rate": 2.494331065759637e-07,
|
| 1148 |
+
"logits/chosen": -2.7794528007507324,
|
| 1149 |
+
"logits/rejected": -2.787205457687378,
|
| 1150 |
+
"logps/chosen": -239.23776245117188,
|
| 1151 |
+
"logps/rejected": -348.8122863769531,
|
| 1152 |
+
"loss": 0.2407,
|
| 1153 |
+
"rewards/accuracies": 0.940625011920929,
|
| 1154 |
+
"rewards/chosen": 1.279539942741394,
|
| 1155 |
+
"rewards/margins": 1.9069591760635376,
|
| 1156 |
+
"rewards/rejected": -0.627419114112854,
|
| 1157 |
+
"step": 810
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.56,
|
| 1161 |
+
"learning_rate": 2.456538170823885e-07,
|
| 1162 |
+
"logits/chosen": -2.7908012866973877,
|
| 1163 |
+
"logits/rejected": -2.775237798690796,
|
| 1164 |
+
"logps/chosen": -237.18807983398438,
|
| 1165 |
+
"logps/rejected": -347.73028564453125,
|
| 1166 |
+
"loss": 0.2346,
|
| 1167 |
+
"rewards/accuracies": 0.938281238079071,
|
| 1168 |
+
"rewards/chosen": 1.2818529605865479,
|
| 1169 |
+
"rewards/margins": 1.891405701637268,
|
| 1170 |
+
"rewards/rejected": -0.6095527410507202,
|
| 1171 |
+
"step": 820
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.56,
|
| 1175 |
+
"learning_rate": 2.418745275888133e-07,
|
| 1176 |
+
"logits/chosen": -2.788677453994751,
|
| 1177 |
+
"logits/rejected": -2.759464740753174,
|
| 1178 |
+
"logps/chosen": -244.3543243408203,
|
| 1179 |
+
"logps/rejected": -384.2773742675781,
|
| 1180 |
+
"loss": 0.249,
|
| 1181 |
+
"rewards/accuracies": 0.9273437261581421,
|
| 1182 |
+
"rewards/chosen": 1.2608978748321533,
|
| 1183 |
+
"rewards/margins": 1.8487341403961182,
|
| 1184 |
+
"rewards/rejected": -0.5878363251686096,
|
| 1185 |
+
"step": 830
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.57,
|
| 1189 |
+
"learning_rate": 2.3809523809523806e-07,
|
| 1190 |
+
"logits/chosen": -2.7865688800811768,
|
| 1191 |
+
"logits/rejected": -2.744267463684082,
|
| 1192 |
+
"logps/chosen": -225.56716918945312,
|
| 1193 |
+
"logps/rejected": -373.64788818359375,
|
| 1194 |
+
"loss": 0.2401,
|
| 1195 |
+
"rewards/accuracies": 0.940625011920929,
|
| 1196 |
+
"rewards/chosen": 1.2590898275375366,
|
| 1197 |
+
"rewards/margins": 1.872513771057129,
|
| 1198 |
+
"rewards/rejected": -0.6134239435195923,
|
| 1199 |
+
"step": 840
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.58,
|
| 1203 |
+
"learning_rate": 2.3431594860166287e-07,
|
| 1204 |
+
"logits/chosen": -2.763679027557373,
|
| 1205 |
+
"logits/rejected": -2.7585010528564453,
|
| 1206 |
+
"logps/chosen": -234.14706420898438,
|
| 1207 |
+
"logps/rejected": -332.43975830078125,
|
| 1208 |
+
"loss": 0.2506,
|
| 1209 |
+
"rewards/accuracies": 0.934374988079071,
|
| 1210 |
+
"rewards/chosen": 1.2429834604263306,
|
| 1211 |
+
"rewards/margins": 1.8476206064224243,
|
| 1212 |
+
"rewards/rejected": -0.6046372056007385,
|
| 1213 |
+
"step": 850
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.58,
|
| 1217 |
+
"learning_rate": 2.3053665910808768e-07,
|
| 1218 |
+
"logits/chosen": -2.7579002380371094,
|
| 1219 |
+
"logits/rejected": -2.7620043754577637,
|
| 1220 |
+
"logps/chosen": -236.3244171142578,
|
| 1221 |
+
"logps/rejected": -339.3128356933594,
|
| 1222 |
+
"loss": 0.2543,
|
| 1223 |
+
"rewards/accuracies": 0.936718761920929,
|
| 1224 |
+
"rewards/chosen": 1.218972086906433,
|
| 1225 |
+
"rewards/margins": 1.8291162252426147,
|
| 1226 |
+
"rewards/rejected": -0.6101440191268921,
|
| 1227 |
+
"step": 860
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.59,
|
| 1231 |
+
"learning_rate": 2.2675736961451246e-07,
|
| 1232 |
+
"logits/chosen": -2.7839019298553467,
|
| 1233 |
+
"logits/rejected": -2.7369167804718018,
|
| 1234 |
+
"logps/chosen": -219.27053833007812,
|
| 1235 |
+
"logps/rejected": -405.5704650878906,
|
| 1236 |
+
"loss": 0.2458,
|
| 1237 |
+
"rewards/accuracies": 0.9312499761581421,
|
| 1238 |
+
"rewards/chosen": 1.2801100015640259,
|
| 1239 |
+
"rewards/margins": 1.8934139013290405,
|
| 1240 |
+
"rewards/rejected": -0.6133038997650146,
|
| 1241 |
+
"step": 870
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.6,
|
| 1245 |
+
"learning_rate": 2.2297808012093725e-07,
|
| 1246 |
+
"logits/chosen": -2.782578945159912,
|
| 1247 |
+
"logits/rejected": -2.7683374881744385,
|
| 1248 |
+
"logps/chosen": -245.6527099609375,
|
| 1249 |
+
"logps/rejected": -378.6884765625,
|
| 1250 |
+
"loss": 0.2384,
|
| 1251 |
+
"rewards/accuracies": 0.934374988079071,
|
| 1252 |
+
"rewards/chosen": 1.321287751197815,
|
| 1253 |
+
"rewards/margins": 1.9386436939239502,
|
| 1254 |
+
"rewards/rejected": -0.6173557043075562,
|
| 1255 |
+
"step": 880
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.61,
|
| 1259 |
+
"learning_rate": 2.1919879062736206e-07,
|
| 1260 |
+
"logits/chosen": -2.7775015830993652,
|
| 1261 |
+
"logits/rejected": -2.752042293548584,
|
| 1262 |
+
"logps/chosen": -229.3787078857422,
|
| 1263 |
+
"logps/rejected": -356.0593566894531,
|
| 1264 |
+
"loss": 0.2423,
|
| 1265 |
+
"rewards/accuracies": 0.934374988079071,
|
| 1266 |
+
"rewards/chosen": 1.2723052501678467,
|
| 1267 |
+
"rewards/margins": 1.9301214218139648,
|
| 1268 |
+
"rewards/rejected": -0.6578160524368286,
|
| 1269 |
+
"step": 890
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.61,
|
| 1273 |
+
"learning_rate": 2.1541950113378684e-07,
|
| 1274 |
+
"logits/chosen": -2.768510580062866,
|
| 1275 |
+
"logits/rejected": -2.7404208183288574,
|
| 1276 |
+
"logps/chosen": -265.3998107910156,
|
| 1277 |
+
"logps/rejected": -373.4928283691406,
|
| 1278 |
+
"loss": 0.2467,
|
| 1279 |
+
"rewards/accuracies": 0.9320312738418579,
|
| 1280 |
+
"rewards/chosen": 1.264615774154663,
|
| 1281 |
+
"rewards/margins": 1.920330286026001,
|
| 1282 |
+
"rewards/rejected": -0.6557145714759827,
|
| 1283 |
+
"step": 900
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.62,
|
| 1287 |
+
"learning_rate": 2.1164021164021162e-07,
|
| 1288 |
+
"logits/chosen": -2.7891170978546143,
|
| 1289 |
+
"logits/rejected": -2.7741641998291016,
|
| 1290 |
+
"logps/chosen": -220.24307250976562,
|
| 1291 |
+
"logps/rejected": -358.5487976074219,
|
| 1292 |
+
"loss": 0.2284,
|
| 1293 |
+
"rewards/accuracies": 0.94140625,
|
| 1294 |
+
"rewards/chosen": 1.304023027420044,
|
| 1295 |
+
"rewards/margins": 1.9829524755477905,
|
| 1296 |
+
"rewards/rejected": -0.6789294481277466,
|
| 1297 |
+
"step": 910
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.63,
|
| 1301 |
+
"learning_rate": 2.0786092214663643e-07,
|
| 1302 |
+
"logits/chosen": -2.7575926780700684,
|
| 1303 |
+
"logits/rejected": -2.7642369270324707,
|
| 1304 |
+
"logps/chosen": -234.12026977539062,
|
| 1305 |
+
"logps/rejected": -384.3020935058594,
|
| 1306 |
+
"loss": 0.2373,
|
| 1307 |
+
"rewards/accuracies": 0.94140625,
|
| 1308 |
+
"rewards/chosen": 1.2832618951797485,
|
| 1309 |
+
"rewards/margins": 1.9688091278076172,
|
| 1310 |
+
"rewards/rejected": -0.6855469942092896,
|
| 1311 |
+
"step": 920
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.63,
|
| 1315 |
+
"learning_rate": 2.0408163265306121e-07,
|
| 1316 |
+
"logits/chosen": -2.766233444213867,
|
| 1317 |
+
"logits/rejected": -2.7951343059539795,
|
| 1318 |
+
"logps/chosen": -244.18026733398438,
|
| 1319 |
+
"logps/rejected": -320.21771240234375,
|
| 1320 |
+
"loss": 0.2259,
|
| 1321 |
+
"rewards/accuracies": 0.9398437738418579,
|
| 1322 |
+
"rewards/chosen": 1.312534213066101,
|
| 1323 |
+
"rewards/margins": 2.0482983589172363,
|
| 1324 |
+
"rewards/rejected": -0.7357643246650696,
|
| 1325 |
+
"step": 930
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.64,
|
| 1329 |
+
"learning_rate": 2.0030234315948602e-07,
|
| 1330 |
+
"logits/chosen": -2.768256664276123,
|
| 1331 |
+
"logits/rejected": -2.7545723915100098,
|
| 1332 |
+
"logps/chosen": -248.816650390625,
|
| 1333 |
+
"logps/rejected": -401.00958251953125,
|
| 1334 |
+
"loss": 0.234,
|
| 1335 |
+
"rewards/accuracies": 0.9320312738418579,
|
| 1336 |
+
"rewards/chosen": 1.3067686557769775,
|
| 1337 |
+
"rewards/margins": 2.003986358642578,
|
| 1338 |
+
"rewards/rejected": -0.6972178816795349,
|
| 1339 |
+
"step": 940
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.65,
|
| 1343 |
+
"learning_rate": 1.965230536659108e-07,
|
| 1344 |
+
"logits/chosen": -2.7718937397003174,
|
| 1345 |
+
"logits/rejected": -2.7864131927490234,
|
| 1346 |
+
"logps/chosen": -245.76220703125,
|
| 1347 |
+
"logps/rejected": -350.4901428222656,
|
| 1348 |
+
"loss": 0.2342,
|
| 1349 |
+
"rewards/accuracies": 0.938281238079071,
|
| 1350 |
+
"rewards/chosen": 1.3374592065811157,
|
| 1351 |
+
"rewards/margins": 2.001889228820801,
|
| 1352 |
+
"rewards/rejected": -0.6644300222396851,
|
| 1353 |
+
"step": 950
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.65,
|
| 1357 |
+
"learning_rate": 1.927437641723356e-07,
|
| 1358 |
+
"logits/chosen": -2.7670133113861084,
|
| 1359 |
+
"logits/rejected": -2.76993465423584,
|
| 1360 |
+
"logps/chosen": -227.41748046875,
|
| 1361 |
+
"logps/rejected": -354.6375427246094,
|
| 1362 |
+
"loss": 0.2386,
|
| 1363 |
+
"rewards/accuracies": 0.92578125,
|
| 1364 |
+
"rewards/chosen": 1.3317902088165283,
|
| 1365 |
+
"rewards/margins": 1.9966375827789307,
|
| 1366 |
+
"rewards/rejected": -0.6648473739624023,
|
| 1367 |
+
"step": 960
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.66,
|
| 1371 |
+
"learning_rate": 1.889644746787604e-07,
|
| 1372 |
+
"logits/chosen": -2.7860965728759766,
|
| 1373 |
+
"logits/rejected": -2.776639699935913,
|
| 1374 |
+
"logps/chosen": -257.2185363769531,
|
| 1375 |
+
"logps/rejected": -302.48846435546875,
|
| 1376 |
+
"loss": 0.2278,
|
| 1377 |
+
"rewards/accuracies": 0.9390624761581421,
|
| 1378 |
+
"rewards/chosen": 1.352912187576294,
|
| 1379 |
+
"rewards/margins": 2.036379814147949,
|
| 1380 |
+
"rewards/rejected": -0.6834677457809448,
|
| 1381 |
+
"step": 970
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.67,
|
| 1385 |
+
"learning_rate": 1.8518518518518516e-07,
|
| 1386 |
+
"logits/chosen": -2.7641091346740723,
|
| 1387 |
+
"logits/rejected": -2.7789313793182373,
|
| 1388 |
+
"logps/chosen": -256.19476318359375,
|
| 1389 |
+
"logps/rejected": -390.69549560546875,
|
| 1390 |
+
"loss": 0.2325,
|
| 1391 |
+
"rewards/accuracies": 0.942187488079071,
|
| 1392 |
+
"rewards/chosen": 1.3011709451675415,
|
| 1393 |
+
"rewards/margins": 2.039425849914551,
|
| 1394 |
+
"rewards/rejected": -0.738254964351654,
|
| 1395 |
+
"step": 980
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.67,
|
| 1399 |
+
"learning_rate": 1.8140589569160996e-07,
|
| 1400 |
+
"logits/chosen": -2.795642614364624,
|
| 1401 |
+
"logits/rejected": -2.7746355533599854,
|
| 1402 |
+
"logps/chosen": -234.4689483642578,
|
| 1403 |
+
"logps/rejected": -395.40618896484375,
|
| 1404 |
+
"loss": 0.223,
|
| 1405 |
+
"rewards/accuracies": 0.9390624761581421,
|
| 1406 |
+
"rewards/chosen": 1.3416879177093506,
|
| 1407 |
+
"rewards/margins": 2.1014368534088135,
|
| 1408 |
+
"rewards/rejected": -0.7597488164901733,
|
| 1409 |
+
"step": 990
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.68,
|
| 1413 |
+
"learning_rate": 1.7762660619803477e-07,
|
| 1414 |
+
"logits/chosen": -2.7756259441375732,
|
| 1415 |
+
"logits/rejected": -2.741664409637451,
|
| 1416 |
+
"logps/chosen": -242.3101348876953,
|
| 1417 |
+
"logps/rejected": -363.46160888671875,
|
| 1418 |
+
"loss": 0.2285,
|
| 1419 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1420 |
+
"rewards/chosen": 1.3446866273880005,
|
| 1421 |
+
"rewards/margins": 2.070406436920166,
|
| 1422 |
+
"rewards/rejected": -0.7257199287414551,
|
| 1423 |
+
"step": 1000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.69,
|
| 1427 |
+
"learning_rate": 1.7384731670445953e-07,
|
| 1428 |
+
"logits/chosen": -2.7595419883728027,
|
| 1429 |
+
"logits/rejected": -2.7858798503875732,
|
| 1430 |
+
"logps/chosen": -259.9520568847656,
|
| 1431 |
+
"logps/rejected": -358.3509216308594,
|
| 1432 |
+
"loss": 0.2273,
|
| 1433 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 1434 |
+
"rewards/chosen": 1.3033568859100342,
|
| 1435 |
+
"rewards/margins": 2.0887067317962646,
|
| 1436 |
+
"rewards/rejected": -0.78534996509552,
|
| 1437 |
+
"step": 1010
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.69,
|
| 1441 |
+
"learning_rate": 1.7006802721088434e-07,
|
| 1442 |
+
"logits/chosen": -2.768449068069458,
|
| 1443 |
+
"logits/rejected": -2.7718656063079834,
|
| 1444 |
+
"logps/chosen": -238.11740112304688,
|
| 1445 |
+
"logps/rejected": -354.0820007324219,
|
| 1446 |
+
"loss": 0.236,
|
| 1447 |
+
"rewards/accuracies": 0.9351562261581421,
|
| 1448 |
+
"rewards/chosen": 1.3048899173736572,
|
| 1449 |
+
"rewards/margins": 2.033565044403076,
|
| 1450 |
+
"rewards/rejected": -0.7286752462387085,
|
| 1451 |
+
"step": 1020
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.7,
|
| 1455 |
+
"learning_rate": 1.6628873771730915e-07,
|
| 1456 |
+
"logits/chosen": -2.7650275230407715,
|
| 1457 |
+
"logits/rejected": -2.7476916313171387,
|
| 1458 |
+
"logps/chosen": -245.41885375976562,
|
| 1459 |
+
"logps/rejected": -343.54437255859375,
|
| 1460 |
+
"loss": 0.2357,
|
| 1461 |
+
"rewards/accuracies": 0.928906261920929,
|
| 1462 |
+
"rewards/chosen": 1.316489815711975,
|
| 1463 |
+
"rewards/margins": 2.062798023223877,
|
| 1464 |
+
"rewards/rejected": -0.7463082671165466,
|
| 1465 |
+
"step": 1030
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.71,
|
| 1469 |
+
"learning_rate": 1.6250944822373393e-07,
|
| 1470 |
+
"logits/chosen": -2.7734358310699463,
|
| 1471 |
+
"logits/rejected": -2.7748751640319824,
|
| 1472 |
+
"logps/chosen": -237.48538208007812,
|
| 1473 |
+
"logps/rejected": -389.1809997558594,
|
| 1474 |
+
"loss": 0.2308,
|
| 1475 |
+
"rewards/accuracies": 0.930468738079071,
|
| 1476 |
+
"rewards/chosen": 1.2910696268081665,
|
| 1477 |
+
"rewards/margins": 2.079051971435547,
|
| 1478 |
+
"rewards/rejected": -0.7879821062088013,
|
| 1479 |
+
"step": 1040
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.71,
|
| 1483 |
+
"learning_rate": 1.5873015873015872e-07,
|
| 1484 |
+
"logits/chosen": -2.7670979499816895,
|
| 1485 |
+
"logits/rejected": -2.769535779953003,
|
| 1486 |
+
"logps/chosen": -214.1968536376953,
|
| 1487 |
+
"logps/rejected": -331.4734802246094,
|
| 1488 |
+
"loss": 0.224,
|
| 1489 |
+
"rewards/accuracies": 0.9390624761581421,
|
| 1490 |
+
"rewards/chosen": 1.377071738243103,
|
| 1491 |
+
"rewards/margins": 2.1104674339294434,
|
| 1492 |
+
"rewards/rejected": -0.7333956956863403,
|
| 1493 |
+
"step": 1050
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.72,
|
| 1497 |
+
"learning_rate": 1.5495086923658353e-07,
|
| 1498 |
+
"logits/chosen": -2.789698839187622,
|
| 1499 |
+
"logits/rejected": -2.7418367862701416,
|
| 1500 |
+
"logps/chosen": -233.3469696044922,
|
| 1501 |
+
"logps/rejected": -372.7334289550781,
|
| 1502 |
+
"loss": 0.2046,
|
| 1503 |
+
"rewards/accuracies": 0.948437511920929,
|
| 1504 |
+
"rewards/chosen": 1.4205210208892822,
|
| 1505 |
+
"rewards/margins": 2.2147347927093506,
|
| 1506 |
+
"rewards/rejected": -0.7942138910293579,
|
| 1507 |
+
"step": 1060
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.73,
|
| 1511 |
+
"learning_rate": 1.511715797430083e-07,
|
| 1512 |
+
"logits/chosen": -2.7732410430908203,
|
| 1513 |
+
"logits/rejected": -2.7837493419647217,
|
| 1514 |
+
"logps/chosen": -240.2108917236328,
|
| 1515 |
+
"logps/rejected": -340.86712646484375,
|
| 1516 |
+
"loss": 0.2229,
|
| 1517 |
+
"rewards/accuracies": 0.932812511920929,
|
| 1518 |
+
"rewards/chosen": 1.365039348602295,
|
| 1519 |
+
"rewards/margins": 2.149728298187256,
|
| 1520 |
+
"rewards/rejected": -0.7846890091896057,
|
| 1521 |
+
"step": 1070
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.73,
|
| 1525 |
+
"learning_rate": 1.473922902494331e-07,
|
| 1526 |
+
"logits/chosen": -2.762357711791992,
|
| 1527 |
+
"logits/rejected": -2.7503538131713867,
|
| 1528 |
+
"logps/chosen": -245.73129272460938,
|
| 1529 |
+
"logps/rejected": -367.2342529296875,
|
| 1530 |
+
"loss": 0.2247,
|
| 1531 |
+
"rewards/accuracies": 0.934374988079071,
|
| 1532 |
+
"rewards/chosen": 1.3420137166976929,
|
| 1533 |
+
"rewards/margins": 2.1435036659240723,
|
| 1534 |
+
"rewards/rejected": -0.8014899492263794,
|
| 1535 |
+
"step": 1080
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.74,
|
| 1539 |
+
"learning_rate": 1.436130007558579e-07,
|
| 1540 |
+
"logits/chosen": -2.786447048187256,
|
| 1541 |
+
"logits/rejected": -2.7433903217315674,
|
| 1542 |
+
"logps/chosen": -259.77923583984375,
|
| 1543 |
+
"logps/rejected": -384.2717590332031,
|
| 1544 |
+
"loss": 0.2176,
|
| 1545 |
+
"rewards/accuracies": 0.938281238079071,
|
| 1546 |
+
"rewards/chosen": 1.4291341304779053,
|
| 1547 |
+
"rewards/margins": 2.1485352516174316,
|
| 1548 |
+
"rewards/rejected": -0.7194010019302368,
|
| 1549 |
+
"step": 1090
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.75,
|
| 1553 |
+
"learning_rate": 1.3983371126228268e-07,
|
| 1554 |
+
"logits/chosen": -2.766045570373535,
|
| 1555 |
+
"logits/rejected": -2.783592700958252,
|
| 1556 |
+
"logps/chosen": -258.433349609375,
|
| 1557 |
+
"logps/rejected": -356.44293212890625,
|
| 1558 |
+
"loss": 0.2166,
|
| 1559 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 1560 |
+
"rewards/chosen": 1.3983967304229736,
|
| 1561 |
+
"rewards/margins": 2.2020390033721924,
|
| 1562 |
+
"rewards/rejected": -0.8036419153213501,
|
| 1563 |
+
"step": 1100
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.75,
|
| 1567 |
+
"learning_rate": 1.360544217687075e-07,
|
| 1568 |
+
"logits/chosen": -2.784245491027832,
|
| 1569 |
+
"logits/rejected": -2.7566187381744385,
|
| 1570 |
+
"logps/chosen": -251.7339324951172,
|
| 1571 |
+
"logps/rejected": -356.1120300292969,
|
| 1572 |
+
"loss": 0.2042,
|
| 1573 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1574 |
+
"rewards/chosen": 1.394803762435913,
|
| 1575 |
+
"rewards/margins": 2.219846487045288,
|
| 1576 |
+
"rewards/rejected": -0.8250430822372437,
|
| 1577 |
+
"step": 1110
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.76,
|
| 1581 |
+
"learning_rate": 1.3227513227513225e-07,
|
| 1582 |
+
"logits/chosen": -2.768209218978882,
|
| 1583 |
+
"logits/rejected": -2.7927510738372803,
|
| 1584 |
+
"logps/chosen": -250.1661376953125,
|
| 1585 |
+
"logps/rejected": -341.25396728515625,
|
| 1586 |
+
"loss": 0.2216,
|
| 1587 |
+
"rewards/accuracies": 0.936718761920929,
|
| 1588 |
+
"rewards/chosen": 1.3784762620925903,
|
| 1589 |
+
"rewards/margins": 2.1875884532928467,
|
| 1590 |
+
"rewards/rejected": -0.8091121912002563,
|
| 1591 |
+
"step": 1120
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.77,
|
| 1595 |
+
"learning_rate": 1.2849584278155706e-07,
|
| 1596 |
+
"logits/chosen": -2.755992889404297,
|
| 1597 |
+
"logits/rejected": -2.7883083820343018,
|
| 1598 |
+
"logps/chosen": -258.3106384277344,
|
| 1599 |
+
"logps/rejected": -338.23822021484375,
|
| 1600 |
+
"loss": 0.2233,
|
| 1601 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 1602 |
+
"rewards/chosen": 1.371977686882019,
|
| 1603 |
+
"rewards/margins": 2.1515755653381348,
|
| 1604 |
+
"rewards/rejected": -0.7795979380607605,
|
| 1605 |
+
"step": 1130
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.77,
|
| 1609 |
+
"learning_rate": 1.2471655328798184e-07,
|
| 1610 |
+
"logits/chosen": -2.765443801879883,
|
| 1611 |
+
"logits/rejected": -2.773919105529785,
|
| 1612 |
+
"logps/chosen": -228.51766967773438,
|
| 1613 |
+
"logps/rejected": -353.0353698730469,
|
| 1614 |
+
"loss": 0.2184,
|
| 1615 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 1616 |
+
"rewards/chosen": 1.3785618543624878,
|
| 1617 |
+
"rewards/margins": 2.173300266265869,
|
| 1618 |
+
"rewards/rejected": -0.7947384119033813,
|
| 1619 |
+
"step": 1140
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.78,
|
| 1623 |
+
"learning_rate": 1.2093726379440665e-07,
|
| 1624 |
+
"logits/chosen": -2.7806408405303955,
|
| 1625 |
+
"logits/rejected": -2.756528854370117,
|
| 1626 |
+
"logps/chosen": -227.71621704101562,
|
| 1627 |
+
"logps/rejected": -391.3194580078125,
|
| 1628 |
+
"loss": 0.2113,
|
| 1629 |
+
"rewards/accuracies": 0.9476562738418579,
|
| 1630 |
+
"rewards/chosen": 1.3853504657745361,
|
| 1631 |
+
"rewards/margins": 2.21071195602417,
|
| 1632 |
+
"rewards/rejected": -0.8253618478775024,
|
| 1633 |
+
"step": 1150
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.79,
|
| 1637 |
+
"learning_rate": 1.1715797430083144e-07,
|
| 1638 |
+
"logits/chosen": -2.7610268592834473,
|
| 1639 |
+
"logits/rejected": -2.7615675926208496,
|
| 1640 |
+
"logps/chosen": -269.79010009765625,
|
| 1641 |
+
"logps/rejected": -378.21209716796875,
|
| 1642 |
+
"loss": 0.2102,
|
| 1643 |
+
"rewards/accuracies": 0.9398437738418579,
|
| 1644 |
+
"rewards/chosen": 1.3920191526412964,
|
| 1645 |
+
"rewards/margins": 2.2798304557800293,
|
| 1646 |
+
"rewards/rejected": -0.887811541557312,
|
| 1647 |
+
"step": 1160
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.8,
|
| 1651 |
+
"learning_rate": 1.1337868480725623e-07,
|
| 1652 |
+
"logits/chosen": -2.776198625564575,
|
| 1653 |
+
"logits/rejected": -2.768550395965576,
|
| 1654 |
+
"logps/chosen": -246.81887817382812,
|
| 1655 |
+
"logps/rejected": -365.49249267578125,
|
| 1656 |
+
"loss": 0.2134,
|
| 1657 |
+
"rewards/accuracies": 0.93359375,
|
| 1658 |
+
"rewards/chosen": 1.4072265625,
|
| 1659 |
+
"rewards/margins": 2.2190985679626465,
|
| 1660 |
+
"rewards/rejected": -0.8118720054626465,
|
| 1661 |
+
"step": 1170
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.8,
|
| 1665 |
+
"learning_rate": 1.0959939531368103e-07,
|
| 1666 |
+
"logits/chosen": -2.782680034637451,
|
| 1667 |
+
"logits/rejected": -2.7389519214630127,
|
| 1668 |
+
"logps/chosen": -239.7198944091797,
|
| 1669 |
+
"logps/rejected": -363.89215087890625,
|
| 1670 |
+
"loss": 0.2147,
|
| 1671 |
+
"rewards/accuracies": 0.942187488079071,
|
| 1672 |
+
"rewards/chosen": 1.3731368780136108,
|
| 1673 |
+
"rewards/margins": 2.227461814880371,
|
| 1674 |
+
"rewards/rejected": -0.8543251156806946,
|
| 1675 |
+
"step": 1180
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.81,
|
| 1679 |
+
"learning_rate": 1.0582010582010581e-07,
|
| 1680 |
+
"logits/chosen": -2.7652835845947266,
|
| 1681 |
+
"logits/rejected": -2.74135160446167,
|
| 1682 |
+
"logps/chosen": -223.03579711914062,
|
| 1683 |
+
"logps/rejected": -407.8848571777344,
|
| 1684 |
+
"loss": 0.2255,
|
| 1685 |
+
"rewards/accuracies": 0.930468738079071,
|
| 1686 |
+
"rewards/chosen": 1.3337465524673462,
|
| 1687 |
+
"rewards/margins": 2.1734132766723633,
|
| 1688 |
+
"rewards/rejected": -0.8396667242050171,
|
| 1689 |
+
"step": 1190
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.82,
|
| 1693 |
+
"learning_rate": 1.0204081632653061e-07,
|
| 1694 |
+
"logits/chosen": -2.785404920578003,
|
| 1695 |
+
"logits/rejected": -2.768556594848633,
|
| 1696 |
+
"logps/chosen": -217.6273651123047,
|
| 1697 |
+
"logps/rejected": -390.0624694824219,
|
| 1698 |
+
"loss": 0.2004,
|
| 1699 |
+
"rewards/accuracies": 0.9515625238418579,
|
| 1700 |
+
"rewards/chosen": 1.452343225479126,
|
| 1701 |
+
"rewards/margins": 2.3132705688476562,
|
| 1702 |
+
"rewards/rejected": -0.8609271049499512,
|
| 1703 |
+
"step": 1200
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.82,
|
| 1707 |
+
"learning_rate": 9.82615268329554e-08,
|
| 1708 |
+
"logits/chosen": -2.7641220092773438,
|
| 1709 |
+
"logits/rejected": -2.7403550148010254,
|
| 1710 |
+
"logps/chosen": -254.1678466796875,
|
| 1711 |
+
"logps/rejected": -381.2986145019531,
|
| 1712 |
+
"loss": 0.2165,
|
| 1713 |
+
"rewards/accuracies": 0.938281238079071,
|
| 1714 |
+
"rewards/chosen": 1.3716920614242554,
|
| 1715 |
+
"rewards/margins": 2.208040952682495,
|
| 1716 |
+
"rewards/rejected": -0.8363490104675293,
|
| 1717 |
+
"step": 1210
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.83,
|
| 1721 |
+
"learning_rate": 9.44822373393802e-08,
|
| 1722 |
+
"logits/chosen": -2.7834503650665283,
|
| 1723 |
+
"logits/rejected": -2.7499313354492188,
|
| 1724 |
+
"logps/chosen": -220.555908203125,
|
| 1725 |
+
"logps/rejected": -353.34454345703125,
|
| 1726 |
+
"loss": 0.2073,
|
| 1727 |
+
"rewards/accuracies": 0.936718761920929,
|
| 1728 |
+
"rewards/chosen": 1.4545724391937256,
|
| 1729 |
+
"rewards/margins": 2.2807674407958984,
|
| 1730 |
+
"rewards/rejected": -0.826195240020752,
|
| 1731 |
+
"step": 1220
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.84,
|
| 1735 |
+
"learning_rate": 9.070294784580498e-08,
|
| 1736 |
+
"logits/chosen": -2.7742843627929688,
|
| 1737 |
+
"logits/rejected": -2.7704269886016846,
|
| 1738 |
+
"logps/chosen": -240.16586303710938,
|
| 1739 |
+
"logps/rejected": -341.08270263671875,
|
| 1740 |
+
"loss": 0.2097,
|
| 1741 |
+
"rewards/accuracies": 0.9398437738418579,
|
| 1742 |
+
"rewards/chosen": 1.3970229625701904,
|
| 1743 |
+
"rewards/margins": 2.205933094024658,
|
| 1744 |
+
"rewards/rejected": -0.8089098930358887,
|
| 1745 |
+
"step": 1230
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.84,
|
| 1749 |
+
"learning_rate": 8.692365835222977e-08,
|
| 1750 |
+
"logits/chosen": -2.7731618881225586,
|
| 1751 |
+
"logits/rejected": -2.7807064056396484,
|
| 1752 |
+
"logps/chosen": -246.8760223388672,
|
| 1753 |
+
"logps/rejected": -394.09661865234375,
|
| 1754 |
+
"loss": 0.1942,
|
| 1755 |
+
"rewards/accuracies": 0.94921875,
|
| 1756 |
+
"rewards/chosen": 1.4174280166625977,
|
| 1757 |
+
"rewards/margins": 2.335336446762085,
|
| 1758 |
+
"rewards/rejected": -0.9179089665412903,
|
| 1759 |
+
"step": 1240
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.85,
|
| 1763 |
+
"learning_rate": 8.314436885865457e-08,
|
| 1764 |
+
"logits/chosen": -2.7794883251190186,
|
| 1765 |
+
"logits/rejected": -2.7599997520446777,
|
| 1766 |
+
"logps/chosen": -234.8397979736328,
|
| 1767 |
+
"logps/rejected": -354.03411865234375,
|
| 1768 |
+
"loss": 0.2101,
|
| 1769 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 1770 |
+
"rewards/chosen": 1.3885688781738281,
|
| 1771 |
+
"rewards/margins": 2.2632603645324707,
|
| 1772 |
+
"rewards/rejected": -0.8746916651725769,
|
| 1773 |
+
"step": 1250
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.86,
|
| 1777 |
+
"learning_rate": 7.936507936507936e-08,
|
| 1778 |
+
"logits/chosen": -2.7606375217437744,
|
| 1779 |
+
"logits/rejected": -2.7535159587860107,
|
| 1780 |
+
"logps/chosen": -225.62606811523438,
|
| 1781 |
+
"logps/rejected": -382.0788269042969,
|
| 1782 |
+
"loss": 0.2247,
|
| 1783 |
+
"rewards/accuracies": 0.9375,
|
| 1784 |
+
"rewards/chosen": 1.3528351783752441,
|
| 1785 |
+
"rewards/margins": 2.18499755859375,
|
| 1786 |
+
"rewards/rejected": -0.8321624994277954,
|
| 1787 |
+
"step": 1260
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.86,
|
| 1791 |
+
"learning_rate": 7.558578987150415e-08,
|
| 1792 |
+
"logits/chosen": -2.7874550819396973,
|
| 1793 |
+
"logits/rejected": -2.7440848350524902,
|
| 1794 |
+
"logps/chosen": -216.8153533935547,
|
| 1795 |
+
"logps/rejected": -372.12982177734375,
|
| 1796 |
+
"loss": 0.2204,
|
| 1797 |
+
"rewards/accuracies": 0.946093738079071,
|
| 1798 |
+
"rewards/chosen": 1.3856043815612793,
|
| 1799 |
+
"rewards/margins": 2.1913902759552,
|
| 1800 |
+
"rewards/rejected": -0.8057858347892761,
|
| 1801 |
+
"step": 1270
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.87,
|
| 1805 |
+
"learning_rate": 7.180650037792895e-08,
|
| 1806 |
+
"logits/chosen": -2.7706284523010254,
|
| 1807 |
+
"logits/rejected": -2.7321717739105225,
|
| 1808 |
+
"logps/chosen": -249.1674041748047,
|
| 1809 |
+
"logps/rejected": -390.70855712890625,
|
| 1810 |
+
"loss": 0.2218,
|
| 1811 |
+
"rewards/accuracies": 0.932812511920929,
|
| 1812 |
+
"rewards/chosen": 1.3266818523406982,
|
| 1813 |
+
"rewards/margins": 2.204909086227417,
|
| 1814 |
+
"rewards/rejected": -0.8782272338867188,
|
| 1815 |
+
"step": 1280
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.88,
|
| 1819 |
+
"learning_rate": 6.802721088435375e-08,
|
| 1820 |
+
"logits/chosen": -2.771331310272217,
|
| 1821 |
+
"logits/rejected": -2.7345921993255615,
|
| 1822 |
+
"logps/chosen": -244.32217407226562,
|
| 1823 |
+
"logps/rejected": -395.6925964355469,
|
| 1824 |
+
"loss": 0.2148,
|
| 1825 |
+
"rewards/accuracies": 0.93359375,
|
| 1826 |
+
"rewards/chosen": 1.379319190979004,
|
| 1827 |
+
"rewards/margins": 2.2045130729675293,
|
| 1828 |
+
"rewards/rejected": -0.8251941800117493,
|
| 1829 |
+
"step": 1290
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.88,
|
| 1833 |
+
"learning_rate": 6.424792139077853e-08,
|
| 1834 |
+
"logits/chosen": -2.781578779220581,
|
| 1835 |
+
"logits/rejected": -2.7588868141174316,
|
| 1836 |
+
"logps/chosen": -234.79800415039062,
|
| 1837 |
+
"logps/rejected": -359.72332763671875,
|
| 1838 |
+
"loss": 0.2047,
|
| 1839 |
+
"rewards/accuracies": 0.9476562738418579,
|
| 1840 |
+
"rewards/chosen": 1.4176688194274902,
|
| 1841 |
+
"rewards/margins": 2.3017234802246094,
|
| 1842 |
+
"rewards/rejected": -0.8840547800064087,
|
| 1843 |
+
"step": 1300
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.89,
|
| 1847 |
+
"learning_rate": 6.046863189720333e-08,
|
| 1848 |
+
"logits/chosen": -2.770113706588745,
|
| 1849 |
+
"logits/rejected": -2.73785662651062,
|
| 1850 |
+
"logps/chosen": -244.0814971923828,
|
| 1851 |
+
"logps/rejected": -370.0007019042969,
|
| 1852 |
+
"loss": 0.2086,
|
| 1853 |
+
"rewards/accuracies": 0.9351562261581421,
|
| 1854 |
+
"rewards/chosen": 1.4492876529693604,
|
| 1855 |
+
"rewards/margins": 2.2907984256744385,
|
| 1856 |
+
"rewards/rejected": -0.8415109515190125,
|
| 1857 |
+
"step": 1310
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.9,
|
| 1861 |
+
"learning_rate": 5.6689342403628116e-08,
|
| 1862 |
+
"logits/chosen": -2.753613233566284,
|
| 1863 |
+
"logits/rejected": -2.7601518630981445,
|
| 1864 |
+
"logps/chosen": -250.3900604248047,
|
| 1865 |
+
"logps/rejected": -360.0616455078125,
|
| 1866 |
+
"loss": 0.2099,
|
| 1867 |
+
"rewards/accuracies": 0.9359375238418579,
|
| 1868 |
+
"rewards/chosen": 1.4489208459854126,
|
| 1869 |
+
"rewards/margins": 2.3246617317199707,
|
| 1870 |
+
"rewards/rejected": -0.8757408261299133,
|
| 1871 |
+
"step": 1320
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.9,
|
| 1875 |
+
"learning_rate": 5.2910052910052905e-08,
|
| 1876 |
+
"logits/chosen": -2.7654261589050293,
|
| 1877 |
+
"logits/rejected": -2.7347397804260254,
|
| 1878 |
+
"logps/chosen": -232.9058837890625,
|
| 1879 |
+
"logps/rejected": -352.3494567871094,
|
| 1880 |
+
"loss": 0.2101,
|
| 1881 |
+
"rewards/accuracies": 0.9398437738418579,
|
| 1882 |
+
"rewards/chosen": 1.4914627075195312,
|
| 1883 |
+
"rewards/margins": 2.3455305099487305,
|
| 1884 |
+
"rewards/rejected": -0.8540679216384888,
|
| 1885 |
+
"step": 1330
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.91,
|
| 1889 |
+
"learning_rate": 4.91307634164777e-08,
|
| 1890 |
+
"logits/chosen": -2.7729830741882324,
|
| 1891 |
+
"logits/rejected": -2.7506096363067627,
|
| 1892 |
+
"logps/chosen": -237.5419158935547,
|
| 1893 |
+
"logps/rejected": -361.7286071777344,
|
| 1894 |
+
"loss": 0.2271,
|
| 1895 |
+
"rewards/accuracies": 0.936718761920929,
|
| 1896 |
+
"rewards/chosen": 1.3387925624847412,
|
| 1897 |
+
"rewards/margins": 2.1680846214294434,
|
| 1898 |
+
"rewards/rejected": -0.8292919397354126,
|
| 1899 |
+
"step": 1340
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.92,
|
| 1903 |
+
"learning_rate": 4.535147392290249e-08,
|
| 1904 |
+
"logits/chosen": -2.758366584777832,
|
| 1905 |
+
"logits/rejected": -2.747448444366455,
|
| 1906 |
+
"logps/chosen": -262.02313232421875,
|
| 1907 |
+
"logps/rejected": -371.6409912109375,
|
| 1908 |
+
"loss": 0.2117,
|
| 1909 |
+
"rewards/accuracies": 0.93359375,
|
| 1910 |
+
"rewards/chosen": 1.4255142211914062,
|
| 1911 |
+
"rewards/margins": 2.2661709785461426,
|
| 1912 |
+
"rewards/rejected": -0.8406568765640259,
|
| 1913 |
+
"step": 1350
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.92,
|
| 1917 |
+
"learning_rate": 4.157218442932729e-08,
|
| 1918 |
+
"logits/chosen": -2.7460246086120605,
|
| 1919 |
+
"logits/rejected": -2.7499794960021973,
|
| 1920 |
+
"logps/chosen": -242.7806396484375,
|
| 1921 |
+
"logps/rejected": -374.79736328125,
|
| 1922 |
+
"loss": 0.2305,
|
| 1923 |
+
"rewards/accuracies": 0.9242187738418579,
|
| 1924 |
+
"rewards/chosen": 1.3290668725967407,
|
| 1925 |
+
"rewards/margins": 2.187917470932007,
|
| 1926 |
+
"rewards/rejected": -0.8588504791259766,
|
| 1927 |
+
"step": 1360
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.93,
|
| 1931 |
+
"learning_rate": 3.779289493575208e-08,
|
| 1932 |
+
"logits/chosen": -2.7681326866149902,
|
| 1933 |
+
"logits/rejected": -2.7562155723571777,
|
| 1934 |
+
"logps/chosen": -220.0043487548828,
|
| 1935 |
+
"logps/rejected": -369.31268310546875,
|
| 1936 |
+
"loss": 0.2015,
|
| 1937 |
+
"rewards/accuracies": 0.944531261920929,
|
| 1938 |
+
"rewards/chosen": 1.4147917032241821,
|
| 1939 |
+
"rewards/margins": 2.347784996032715,
|
| 1940 |
+
"rewards/rejected": -0.9329932332038879,
|
| 1941 |
+
"step": 1370
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.94,
|
| 1945 |
+
"learning_rate": 3.4013605442176873e-08,
|
| 1946 |
+
"logits/chosen": -2.7685980796813965,
|
| 1947 |
+
"logits/rejected": -2.761018753051758,
|
| 1948 |
+
"logps/chosen": -244.3848114013672,
|
| 1949 |
+
"logps/rejected": -352.2154235839844,
|
| 1950 |
+
"loss": 0.2147,
|
| 1951 |
+
"rewards/accuracies": 0.9398437738418579,
|
| 1952 |
+
"rewards/chosen": 1.3917274475097656,
|
| 1953 |
+
"rewards/margins": 2.2305819988250732,
|
| 1954 |
+
"rewards/rejected": -0.8388546109199524,
|
| 1955 |
+
"step": 1380
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.94,
|
| 1959 |
+
"learning_rate": 3.023431594860166e-08,
|
| 1960 |
+
"logits/chosen": -2.7724173069000244,
|
| 1961 |
+
"logits/rejected": -2.773851156234741,
|
| 1962 |
+
"logps/chosen": -251.663330078125,
|
| 1963 |
+
"logps/rejected": -341.803466796875,
|
| 1964 |
+
"loss": 0.1992,
|
| 1965 |
+
"rewards/accuracies": 0.94921875,
|
| 1966 |
+
"rewards/chosen": 1.4376652240753174,
|
| 1967 |
+
"rewards/margins": 2.324432134628296,
|
| 1968 |
+
"rewards/rejected": -0.886766791343689,
|
| 1969 |
+
"step": 1390
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.95,
|
| 1973 |
+
"learning_rate": 2.6455026455026453e-08,
|
| 1974 |
+
"logits/chosen": -2.758798122406006,
|
| 1975 |
+
"logits/rejected": -2.763350009918213,
|
| 1976 |
+
"logps/chosen": -238.17745971679688,
|
| 1977 |
+
"logps/rejected": -398.58135986328125,
|
| 1978 |
+
"loss": 0.2004,
|
| 1979 |
+
"rewards/accuracies": 0.9476562738418579,
|
| 1980 |
+
"rewards/chosen": 1.4174001216888428,
|
| 1981 |
+
"rewards/margins": 2.3445682525634766,
|
| 1982 |
+
"rewards/rejected": -0.9271681904792786,
|
| 1983 |
+
"step": 1400
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.96,
|
| 1987 |
+
"learning_rate": 2.2675736961451246e-08,
|
| 1988 |
+
"logits/chosen": -2.7801098823547363,
|
| 1989 |
+
"logits/rejected": -2.7490382194519043,
|
| 1990 |
+
"logps/chosen": -242.81613159179688,
|
| 1991 |
+
"logps/rejected": -361.264892578125,
|
| 1992 |
+
"loss": 0.2077,
|
| 1993 |
+
"rewards/accuracies": 0.94140625,
|
| 1994 |
+
"rewards/chosen": 1.4166629314422607,
|
| 1995 |
+
"rewards/margins": 2.316483736038208,
|
| 1996 |
+
"rewards/rejected": -0.8998208045959473,
|
| 1997 |
+
"step": 1410
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.97,
|
| 2001 |
+
"learning_rate": 1.889644746787604e-08,
|
| 2002 |
+
"logits/chosen": -2.75722336769104,
|
| 2003 |
+
"logits/rejected": -2.7228329181671143,
|
| 2004 |
+
"logps/chosen": -251.5331268310547,
|
| 2005 |
+
"logps/rejected": -375.8110046386719,
|
| 2006 |
+
"loss": 0.2226,
|
| 2007 |
+
"rewards/accuracies": 0.940625011920929,
|
| 2008 |
+
"rewards/chosen": 1.4032243490219116,
|
| 2009 |
+
"rewards/margins": 2.2185873985290527,
|
| 2010 |
+
"rewards/rejected": -0.8153629302978516,
|
| 2011 |
+
"step": 1420
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.97,
|
| 2015 |
+
"learning_rate": 1.511715797430083e-08,
|
| 2016 |
+
"logits/chosen": -2.7710134983062744,
|
| 2017 |
+
"logits/rejected": -2.787081241607666,
|
| 2018 |
+
"logps/chosen": -241.9620361328125,
|
| 2019 |
+
"logps/rejected": -356.4383544921875,
|
| 2020 |
+
"loss": 0.2074,
|
| 2021 |
+
"rewards/accuracies": 0.938281238079071,
|
| 2022 |
+
"rewards/chosen": 1.4058793783187866,
|
| 2023 |
+
"rewards/margins": 2.3387274742126465,
|
| 2024 |
+
"rewards/rejected": -0.9328481554985046,
|
| 2025 |
+
"step": 1430
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.98,
|
| 2029 |
+
"learning_rate": 1.1337868480725623e-08,
|
| 2030 |
+
"logits/chosen": -2.788255214691162,
|
| 2031 |
+
"logits/rejected": -2.790001392364502,
|
| 2032 |
+
"logps/chosen": -249.0662078857422,
|
| 2033 |
+
"logps/rejected": -375.603759765625,
|
| 2034 |
+
"loss": 0.1976,
|
| 2035 |
+
"rewards/accuracies": 0.9429687261581421,
|
| 2036 |
+
"rewards/chosen": 1.4689807891845703,
|
| 2037 |
+
"rewards/margins": 2.375899076461792,
|
| 2038 |
+
"rewards/rejected": -0.9069182276725769,
|
| 2039 |
+
"step": 1440
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.99,
|
| 2043 |
+
"learning_rate": 7.558578987150416e-09,
|
| 2044 |
+
"logits/chosen": -2.762585401535034,
|
| 2045 |
+
"logits/rejected": -2.7085330486297607,
|
| 2046 |
+
"logps/chosen": -238.41751098632812,
|
| 2047 |
+
"logps/rejected": -380.84942626953125,
|
| 2048 |
+
"loss": 0.2228,
|
| 2049 |
+
"rewards/accuracies": 0.936718761920929,
|
| 2050 |
+
"rewards/chosen": 1.4105838537216187,
|
| 2051 |
+
"rewards/margins": 2.214503288269043,
|
| 2052 |
+
"rewards/rejected": -0.8039194345474243,
|
| 2053 |
+
"step": 1450
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.99,
|
| 2057 |
+
"learning_rate": 3.779289493575208e-09,
|
| 2058 |
+
"logits/chosen": -2.7654013633728027,
|
| 2059 |
+
"logits/rejected": -2.7555670738220215,
|
| 2060 |
+
"logps/chosen": -237.80899047851562,
|
| 2061 |
+
"logps/rejected": -345.7412109375,
|
| 2062 |
+
"loss": 0.2026,
|
| 2063 |
+
"rewards/accuracies": 0.9515625238418579,
|
| 2064 |
+
"rewards/chosen": 1.418304443359375,
|
| 2065 |
+
"rewards/margins": 2.327260971069336,
|
| 2066 |
+
"rewards/rejected": -0.9089563488960266,
|
| 2067 |
+
"step": 1460
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 1.0,
|
| 2071 |
+
"learning_rate": 0.0,
|
| 2072 |
+
"logits/chosen": -2.75099515914917,
|
| 2073 |
+
"logits/rejected": -2.7724432945251465,
|
| 2074 |
+
"logps/chosen": -255.767578125,
|
| 2075 |
+
"logps/rejected": -360.5829772949219,
|
| 2076 |
+
"loss": 0.2019,
|
| 2077 |
+
"rewards/accuracies": 0.946093738079071,
|
| 2078 |
+
"rewards/chosen": 1.4355896711349487,
|
| 2079 |
+
"rewards/margins": 2.3439955711364746,
|
| 2080 |
+
"rewards/rejected": -0.9084057807922363,
|
| 2081 |
+
"step": 1470
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 1.0,
|
| 2085 |
+
"eval_logits/chosen": -2.617767333984375,
|
| 2086 |
+
"eval_logits/rejected": -2.721874952316284,
|
| 2087 |
+
"eval_logps/chosen": -238.54788208007812,
|
| 2088 |
+
"eval_logps/rejected": -388.59033203125,
|
| 2089 |
+
"eval_loss": 0.20815864205360413,
|
| 2090 |
+
"eval_rewards/accuracies": 0.9413930773735046,
|
| 2091 |
+
"eval_rewards/chosen": 1.3856867551803589,
|
| 2092 |
+
"eval_rewards/margins": 2.292266845703125,
|
| 2093 |
+
"eval_rewards/rejected": -0.9065799117088318,
|
| 2094 |
+
"eval_runtime": 2798.4996,
|
| 2095 |
+
"eval_samples_per_second": 3.395,
|
| 2096 |
+
"eval_steps_per_second": 0.425,
|
| 2097 |
+
"step": 1470
|
| 2098 |
+
},
|
| 2099 |
+
{
|
| 2100 |
+
"epoch": 1.0,
|
| 2101 |
+
"step": 1470,
|
| 2102 |
+
"total_flos": 0.0,
|
| 2103 |
+
"train_loss": 0.33413780781687524,
|
| 2104 |
+
"train_runtime": 91396.7242,
|
| 2105 |
+
"train_samples_per_second": 2.06,
|
| 2106 |
+
"train_steps_per_second": 0.016
|
| 2107 |
+
}
|
| 2108 |
+
],
|
| 2109 |
+
"logging_steps": 10,
|
| 2110 |
+
"max_steps": 1470,
|
| 2111 |
+
"num_train_epochs": 1,
|
| 2112 |
+
"save_steps": 500,
|
| 2113 |
+
"total_flos": 0.0,
|
| 2114 |
+
"trial_name": null,
|
| 2115 |
+
"trial_params": null
|
| 2116 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:219972c2a6a222c6eefdfc187b05a639de3b1f90c7026b56decc2c8bb45f034c
|
| 3 |
+
size 4728
|