Model save
Browse files- README.md +73 -0
- added_tokens.json +13 -0
- all_results.json +9 -0
- config.json +140 -0
- generation_config.json +11 -0
- model-00001-of-00002.safetensors +3 -0
- model-00002-of-00002.safetensors +3 -0
- model.safetensors.index.json +202 -0
- runs/Jun21_16-34-46_131-186-38-174/events.out.tfevents.1718988272.131-186-38-174.27809.0 +3 -0
- special_tokens_map.json +30 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +130 -0
- train_results.json +9 -0
- trainer_state.json +1468 -0
- training_args.bin +3 -0
README.md
ADDED
@@ -0,0 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: mit
|
3 |
+
base_model: microsoft/Phi-3-mini-128k-instruct
|
4 |
+
tags:
|
5 |
+
- trl
|
6 |
+
- dpo
|
7 |
+
- generated_from_trainer
|
8 |
+
model-index:
|
9 |
+
- name: phi-3-mini-128k-instruct-simpo-lr-5e-07-gamma-1.5
|
10 |
+
results: []
|
11 |
+
---
|
12 |
+
|
13 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
14 |
+
should probably proofread and complete it, then remove this comment. -->
|
15 |
+
|
16 |
+
# phi-3-mini-128k-instruct-simpo-lr-5e-07-gamma-1.5
|
17 |
+
|
18 |
+
This model is a fine-tuned version of [microsoft/Phi-3-mini-128k-instruct](https://huggingface.co/microsoft/Phi-3-mini-128k-instruct) on an unknown dataset.
|
19 |
+
It achieves the following results on the evaluation set:
|
20 |
+
- Loss: 1.6236
|
21 |
+
- Rewards/chosen: -2.2390
|
22 |
+
- Rewards/rejected: -2.6457
|
23 |
+
- Rewards/accuracies: 0.6210
|
24 |
+
- Rewards/margins: 0.4067
|
25 |
+
- Logps/rejected: -1.0583
|
26 |
+
- Logps/chosen: -0.8956
|
27 |
+
- Logits/rejected: 2.0190
|
28 |
+
- Logits/chosen: 2.0146
|
29 |
+
|
30 |
+
## Model description
|
31 |
+
|
32 |
+
More information needed
|
33 |
+
|
34 |
+
## Intended uses & limitations
|
35 |
+
|
36 |
+
More information needed
|
37 |
+
|
38 |
+
## Training and evaluation data
|
39 |
+
|
40 |
+
More information needed
|
41 |
+
|
42 |
+
## Training procedure
|
43 |
+
|
44 |
+
### Training hyperparameters
|
45 |
+
|
46 |
+
The following hyperparameters were used during training:
|
47 |
+
- learning_rate: 5e-07
|
48 |
+
- train_batch_size: 2
|
49 |
+
- eval_batch_size: 4
|
50 |
+
- seed: 42
|
51 |
+
- distributed_type: multi-GPU
|
52 |
+
- num_devices: 8
|
53 |
+
- gradient_accumulation_steps: 8
|
54 |
+
- total_train_batch_size: 128
|
55 |
+
- total_eval_batch_size: 32
|
56 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
57 |
+
- lr_scheduler_type: cosine
|
58 |
+
- lr_scheduler_warmup_ratio: 0.1
|
59 |
+
- num_epochs: 1
|
60 |
+
|
61 |
+
### Training results
|
62 |
+
|
63 |
+
| Training Loss | Epoch | Step | Validation Loss | Rewards/chosen | Rewards/rejected | Rewards/accuracies | Rewards/margins | Logps/rejected | Logps/chosen | Logits/rejected | Logits/chosen |
|
64 |
+
|:-------------:|:------:|:----:|:---------------:|:--------------:|:----------------:|:------------------:|:---------------:|:--------------:|:------------:|:---------------:|:-------------:|
|
65 |
+
| 1.6417 | 0.8549 | 400 | 1.6236 | -2.2390 | -2.6457 | 0.6210 | 0.4067 | -1.0583 | -0.8956 | 2.0190 | 2.0146 |
|
66 |
+
|
67 |
+
|
68 |
+
### Framework versions
|
69 |
+
|
70 |
+
- Transformers 4.41.2
|
71 |
+
- Pytorch 2.3.1+cu121
|
72 |
+
- Datasets 2.20.0
|
73 |
+
- Tokenizers 0.19.1
|
added_tokens.json
ADDED
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"<|assistant|>": 32001,
|
3 |
+
"<|endoftext|>": 32000,
|
4 |
+
"<|end|>": 32007,
|
5 |
+
"<|placeholder1|>": 32002,
|
6 |
+
"<|placeholder2|>": 32003,
|
7 |
+
"<|placeholder3|>": 32004,
|
8 |
+
"<|placeholder4|>": 32005,
|
9 |
+
"<|placeholder5|>": 32008,
|
10 |
+
"<|placeholder6|>": 32009,
|
11 |
+
"<|system|>": 32006,
|
12 |
+
"<|user|>": 32010
|
13 |
+
}
|
all_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 0.9981298423724285,
|
3 |
+
"total_flos": 0.0,
|
4 |
+
"train_loss": 1.6566840494470372,
|
5 |
+
"train_runtime": 5417.7619,
|
6 |
+
"train_samples": 59876,
|
7 |
+
"train_samples_per_second": 11.052,
|
8 |
+
"train_steps_per_second": 0.086
|
9 |
+
}
|
config.json
ADDED
@@ -0,0 +1,140 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/Phi-3-mini-128k-instruct",
|
3 |
+
"architectures": [
|
4 |
+
"Phi3ForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_bias": false,
|
7 |
+
"attention_dropout": 0.0,
|
8 |
+
"auto_map": {
|
9 |
+
"AutoConfig": "microsoft/Phi-3-mini-128k-instruct--configuration_phi3.Phi3Config",
|
10 |
+
"AutoModelForCausalLM": "microsoft/Phi-3-mini-128k-instruct--modeling_phi3.Phi3ForCausalLM",
|
11 |
+
"AutoModelForSequenceClassification": "microsoft/Phi-3-mini-128k-instruct--modeling_phi3.Phi3ForSequenceClassification",
|
12 |
+
"AutoModelForTokenClassification": "microsoft/Phi-3-mini-128k-instruct--modeling_phi3.Phi3ForTokenClassification"
|
13 |
+
},
|
14 |
+
"bos_token_id": 1,
|
15 |
+
"embd_pdrop": 0.0,
|
16 |
+
"eos_token_id": 32000,
|
17 |
+
"hidden_act": "silu",
|
18 |
+
"hidden_size": 3072,
|
19 |
+
"initializer_range": 0.02,
|
20 |
+
"intermediate_size": 8192,
|
21 |
+
"max_position_embeddings": 131072,
|
22 |
+
"model_type": "phi3",
|
23 |
+
"num_attention_heads": 32,
|
24 |
+
"num_hidden_layers": 32,
|
25 |
+
"num_key_value_heads": 32,
|
26 |
+
"original_max_position_embeddings": 4096,
|
27 |
+
"pad_token_id": 32000,
|
28 |
+
"resid_pdrop": 0.0,
|
29 |
+
"rms_norm_eps": 1e-05,
|
30 |
+
"rope_scaling": {
|
31 |
+
"long_factor": [
|
32 |
+
1.0299999713897705,
|
33 |
+
1.0499999523162842,
|
34 |
+
1.0499999523162842,
|
35 |
+
1.0799999237060547,
|
36 |
+
1.2299998998641968,
|
37 |
+
1.2299998998641968,
|
38 |
+
1.2999999523162842,
|
39 |
+
1.4499999284744263,
|
40 |
+
1.5999999046325684,
|
41 |
+
1.6499998569488525,
|
42 |
+
1.8999998569488525,
|
43 |
+
2.859999895095825,
|
44 |
+
3.68999981880188,
|
45 |
+
5.419999599456787,
|
46 |
+
5.489999771118164,
|
47 |
+
5.489999771118164,
|
48 |
+
9.09000015258789,
|
49 |
+
11.579999923706055,
|
50 |
+
15.65999984741211,
|
51 |
+
15.769999504089355,
|
52 |
+
15.789999961853027,
|
53 |
+
18.360000610351562,
|
54 |
+
21.989999771118164,
|
55 |
+
23.079999923706055,
|
56 |
+
30.009998321533203,
|
57 |
+
32.35000228881836,
|
58 |
+
32.590003967285156,
|
59 |
+
35.56000518798828,
|
60 |
+
39.95000457763672,
|
61 |
+
53.840003967285156,
|
62 |
+
56.20000457763672,
|
63 |
+
57.95000457763672,
|
64 |
+
59.29000473022461,
|
65 |
+
59.77000427246094,
|
66 |
+
59.920005798339844,
|
67 |
+
61.190006256103516,
|
68 |
+
61.96000671386719,
|
69 |
+
62.50000762939453,
|
70 |
+
63.3700065612793,
|
71 |
+
63.48000717163086,
|
72 |
+
63.48000717163086,
|
73 |
+
63.66000747680664,
|
74 |
+
63.850006103515625,
|
75 |
+
64.08000946044922,
|
76 |
+
64.760009765625,
|
77 |
+
64.80001068115234,
|
78 |
+
64.81001281738281,
|
79 |
+
64.81001281738281
|
80 |
+
],
|
81 |
+
"short_factor": [
|
82 |
+
1.05,
|
83 |
+
1.05,
|
84 |
+
1.05,
|
85 |
+
1.1,
|
86 |
+
1.1,
|
87 |
+
1.1500000000000001,
|
88 |
+
1.2000000000000002,
|
89 |
+
1.2500000000000002,
|
90 |
+
1.3000000000000003,
|
91 |
+
1.3500000000000003,
|
92 |
+
1.5000000000000004,
|
93 |
+
2.000000000000001,
|
94 |
+
2.000000000000001,
|
95 |
+
2.000000000000001,
|
96 |
+
2.000000000000001,
|
97 |
+
2.000000000000001,
|
98 |
+
2.000000000000001,
|
99 |
+
2.000000000000001,
|
100 |
+
2.000000000000001,
|
101 |
+
2.000000000000001,
|
102 |
+
2.000000000000001,
|
103 |
+
2.000000000000001,
|
104 |
+
2.000000000000001,
|
105 |
+
2.000000000000001,
|
106 |
+
2.000000000000001,
|
107 |
+
2.000000000000001,
|
108 |
+
2.000000000000001,
|
109 |
+
2.000000000000001,
|
110 |
+
2.000000000000001,
|
111 |
+
2.000000000000001,
|
112 |
+
2.000000000000001,
|
113 |
+
2.000000000000001,
|
114 |
+
2.0500000000000007,
|
115 |
+
2.0500000000000007,
|
116 |
+
2.0500000000000007,
|
117 |
+
2.1000000000000005,
|
118 |
+
2.1000000000000005,
|
119 |
+
2.1000000000000005,
|
120 |
+
2.1500000000000004,
|
121 |
+
2.1500000000000004,
|
122 |
+
2.3499999999999996,
|
123 |
+
2.549999999999999,
|
124 |
+
2.5999999999999988,
|
125 |
+
2.5999999999999988,
|
126 |
+
2.7499999999999982,
|
127 |
+
2.849999999999998,
|
128 |
+
2.849999999999998,
|
129 |
+
2.9499999999999975
|
130 |
+
],
|
131 |
+
"type": "su"
|
132 |
+
},
|
133 |
+
"rope_theta": 10000.0,
|
134 |
+
"sliding_window": 262144,
|
135 |
+
"tie_word_embeddings": false,
|
136 |
+
"torch_dtype": "bfloat16",
|
137 |
+
"transformers_version": "4.41.2",
|
138 |
+
"use_cache": false,
|
139 |
+
"vocab_size": 32064
|
140 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,11 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 1,
|
4 |
+
"eos_token_id": [
|
5 |
+
32000,
|
6 |
+
32001,
|
7 |
+
32007
|
8 |
+
],
|
9 |
+
"pad_token_id": 32000,
|
10 |
+
"transformers_version": "4.41.2"
|
11 |
+
}
|
model-00001-of-00002.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d5e43f01963969b46efa3bb114a99bfe20876daf51668f39d7e18ad4288d2a8d
|
3 |
+
size 4972489328
|
model-00002-of-00002.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1ac11a3fc0a84be8028c27a527073d4bcb64a3884def72eaa84043491956c8f3
|
3 |
+
size 2669692552
|
model.safetensors.index.json
ADDED
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"metadata": {
|
3 |
+
"total_size": 7642159104
|
4 |
+
},
|
5 |
+
"weight_map": {
|
6 |
+
"lm_head.weight": "model-00002-of-00002.safetensors",
|
7 |
+
"model.embed_tokens.weight": "model-00001-of-00002.safetensors",
|
8 |
+
"model.layers.0.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
9 |
+
"model.layers.0.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
10 |
+
"model.layers.0.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
11 |
+
"model.layers.0.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
12 |
+
"model.layers.0.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
13 |
+
"model.layers.0.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
14 |
+
"model.layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
15 |
+
"model.layers.1.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
16 |
+
"model.layers.1.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
17 |
+
"model.layers.1.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
18 |
+
"model.layers.1.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
19 |
+
"model.layers.1.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
20 |
+
"model.layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
21 |
+
"model.layers.10.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
22 |
+
"model.layers.10.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
23 |
+
"model.layers.10.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
24 |
+
"model.layers.10.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
25 |
+
"model.layers.10.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
26 |
+
"model.layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
27 |
+
"model.layers.11.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
28 |
+
"model.layers.11.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
29 |
+
"model.layers.11.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
30 |
+
"model.layers.11.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
31 |
+
"model.layers.11.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
32 |
+
"model.layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
33 |
+
"model.layers.12.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
34 |
+
"model.layers.12.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
35 |
+
"model.layers.12.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
36 |
+
"model.layers.12.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
37 |
+
"model.layers.12.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
38 |
+
"model.layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
39 |
+
"model.layers.13.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
40 |
+
"model.layers.13.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
41 |
+
"model.layers.13.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
42 |
+
"model.layers.13.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
43 |
+
"model.layers.13.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
44 |
+
"model.layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
45 |
+
"model.layers.14.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
46 |
+
"model.layers.14.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
47 |
+
"model.layers.14.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
48 |
+
"model.layers.14.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
49 |
+
"model.layers.14.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
50 |
+
"model.layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
51 |
+
"model.layers.15.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
52 |
+
"model.layers.15.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
53 |
+
"model.layers.15.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
54 |
+
"model.layers.15.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
55 |
+
"model.layers.15.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
56 |
+
"model.layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
57 |
+
"model.layers.16.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
58 |
+
"model.layers.16.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
59 |
+
"model.layers.16.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
60 |
+
"model.layers.16.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
61 |
+
"model.layers.16.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
62 |
+
"model.layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
63 |
+
"model.layers.17.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
64 |
+
"model.layers.17.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
65 |
+
"model.layers.17.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
66 |
+
"model.layers.17.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
67 |
+
"model.layers.17.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
68 |
+
"model.layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
69 |
+
"model.layers.18.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
70 |
+
"model.layers.18.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
71 |
+
"model.layers.18.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
72 |
+
"model.layers.18.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
73 |
+
"model.layers.18.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
74 |
+
"model.layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
75 |
+
"model.layers.19.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
76 |
+
"model.layers.19.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
77 |
+
"model.layers.19.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
78 |
+
"model.layers.19.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
79 |
+
"model.layers.19.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
80 |
+
"model.layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
81 |
+
"model.layers.2.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
82 |
+
"model.layers.2.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
83 |
+
"model.layers.2.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
84 |
+
"model.layers.2.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
85 |
+
"model.layers.2.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
86 |
+
"model.layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
87 |
+
"model.layers.20.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
88 |
+
"model.layers.20.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
89 |
+
"model.layers.20.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
90 |
+
"model.layers.20.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
91 |
+
"model.layers.20.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
92 |
+
"model.layers.21.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
93 |
+
"model.layers.21.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
94 |
+
"model.layers.21.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
95 |
+
"model.layers.21.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
96 |
+
"model.layers.21.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
97 |
+
"model.layers.21.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
98 |
+
"model.layers.22.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
99 |
+
"model.layers.22.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
100 |
+
"model.layers.22.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
101 |
+
"model.layers.22.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
102 |
+
"model.layers.22.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
103 |
+
"model.layers.22.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
104 |
+
"model.layers.23.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
105 |
+
"model.layers.23.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
106 |
+
"model.layers.23.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
107 |
+
"model.layers.23.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
108 |
+
"model.layers.23.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
109 |
+
"model.layers.23.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
110 |
+
"model.layers.24.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
111 |
+
"model.layers.24.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
112 |
+
"model.layers.24.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
113 |
+
"model.layers.24.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
114 |
+
"model.layers.24.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
115 |
+
"model.layers.24.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
116 |
+
"model.layers.25.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
117 |
+
"model.layers.25.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
118 |
+
"model.layers.25.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
119 |
+
"model.layers.25.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
120 |
+
"model.layers.25.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
121 |
+
"model.layers.25.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
122 |
+
"model.layers.26.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
123 |
+
"model.layers.26.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
124 |
+
"model.layers.26.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
125 |
+
"model.layers.26.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
126 |
+
"model.layers.26.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
127 |
+
"model.layers.26.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
128 |
+
"model.layers.27.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
129 |
+
"model.layers.27.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
130 |
+
"model.layers.27.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
131 |
+
"model.layers.27.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
132 |
+
"model.layers.27.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
133 |
+
"model.layers.27.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
134 |
+
"model.layers.28.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
135 |
+
"model.layers.28.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
136 |
+
"model.layers.28.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
137 |
+
"model.layers.28.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
138 |
+
"model.layers.28.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
139 |
+
"model.layers.28.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
140 |
+
"model.layers.29.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
141 |
+
"model.layers.29.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
142 |
+
"model.layers.29.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
143 |
+
"model.layers.29.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
144 |
+
"model.layers.29.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
145 |
+
"model.layers.29.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
146 |
+
"model.layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
147 |
+
"model.layers.3.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
148 |
+
"model.layers.3.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
149 |
+
"model.layers.3.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
150 |
+
"model.layers.3.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
151 |
+
"model.layers.3.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
152 |
+
"model.layers.30.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
153 |
+
"model.layers.30.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
154 |
+
"model.layers.30.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
155 |
+
"model.layers.30.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
156 |
+
"model.layers.30.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
157 |
+
"model.layers.30.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
158 |
+
"model.layers.31.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
159 |
+
"model.layers.31.mlp.down_proj.weight": "model-00002-of-00002.safetensors",
|
160 |
+
"model.layers.31.mlp.gate_up_proj.weight": "model-00002-of-00002.safetensors",
|
161 |
+
"model.layers.31.post_attention_layernorm.weight": "model-00002-of-00002.safetensors",
|
162 |
+
"model.layers.31.self_attn.o_proj.weight": "model-00002-of-00002.safetensors",
|
163 |
+
"model.layers.31.self_attn.qkv_proj.weight": "model-00002-of-00002.safetensors",
|
164 |
+
"model.layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
165 |
+
"model.layers.4.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
166 |
+
"model.layers.4.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
167 |
+
"model.layers.4.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
168 |
+
"model.layers.4.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
169 |
+
"model.layers.4.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
170 |
+
"model.layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
171 |
+
"model.layers.5.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
172 |
+
"model.layers.5.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
173 |
+
"model.layers.5.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
174 |
+
"model.layers.5.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
175 |
+
"model.layers.5.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
176 |
+
"model.layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
177 |
+
"model.layers.6.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
178 |
+
"model.layers.6.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
179 |
+
"model.layers.6.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
180 |
+
"model.layers.6.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
181 |
+
"model.layers.6.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
182 |
+
"model.layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
183 |
+
"model.layers.7.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
184 |
+
"model.layers.7.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
185 |
+
"model.layers.7.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
186 |
+
"model.layers.7.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
187 |
+
"model.layers.7.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
188 |
+
"model.layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
189 |
+
"model.layers.8.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
190 |
+
"model.layers.8.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
191 |
+
"model.layers.8.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
192 |
+
"model.layers.8.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
193 |
+
"model.layers.8.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
194 |
+
"model.layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
195 |
+
"model.layers.9.mlp.down_proj.weight": "model-00001-of-00002.safetensors",
|
196 |
+
"model.layers.9.mlp.gate_up_proj.weight": "model-00001-of-00002.safetensors",
|
197 |
+
"model.layers.9.post_attention_layernorm.weight": "model-00001-of-00002.safetensors",
|
198 |
+
"model.layers.9.self_attn.o_proj.weight": "model-00001-of-00002.safetensors",
|
199 |
+
"model.layers.9.self_attn.qkv_proj.weight": "model-00001-of-00002.safetensors",
|
200 |
+
"model.norm.weight": "model-00002-of-00002.safetensors"
|
201 |
+
}
|
202 |
+
}
|
runs/Jun21_16-34-46_131-186-38-174/events.out.tfevents.1718988272.131-186-38-174.27809.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d7d0e76205db49909994d392502c92b444b4f2a5399db0d71639e5e88863b958
|
3 |
+
size 73664
|
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": "<|endoftext|>",
|
11 |
+
"lstrip": false,
|
12 |
+
"normalized": false,
|
13 |
+
"rstrip": false,
|
14 |
+
"single_word": false
|
15 |
+
},
|
16 |
+
"pad_token": {
|
17 |
+
"content": "<|endoftext|>",
|
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.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
3 |
+
size 499723
|
tokenizer_config.json
ADDED
@@ -0,0 +1,130 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_bos_token": true,
|
3 |
+
"add_eos_token": false,
|
4 |
+
"added_tokens_decoder": {
|
5 |
+
"0": {
|
6 |
+
"content": "<unk>",
|
7 |
+
"lstrip": false,
|
8 |
+
"normalized": false,
|
9 |
+
"rstrip": false,
|
10 |
+
"single_word": false,
|
11 |
+
"special": true
|
12 |
+
},
|
13 |
+
"1": {
|
14 |
+
"content": "<s>",
|
15 |
+
"lstrip": false,
|
16 |
+
"normalized": false,
|
17 |
+
"rstrip": false,
|
18 |
+
"single_word": false,
|
19 |
+
"special": true
|
20 |
+
},
|
21 |
+
"2": {
|
22 |
+
"content": "</s>",
|
23 |
+
"lstrip": false,
|
24 |
+
"normalized": false,
|
25 |
+
"rstrip": true,
|
26 |
+
"single_word": false,
|
27 |
+
"special": false
|
28 |
+
},
|
29 |
+
"32000": {
|
30 |
+
"content": "<|endoftext|>",
|
31 |
+
"lstrip": false,
|
32 |
+
"normalized": false,
|
33 |
+
"rstrip": false,
|
34 |
+
"single_word": false,
|
35 |
+
"special": true
|
36 |
+
},
|
37 |
+
"32001": {
|
38 |
+
"content": "<|assistant|>",
|
39 |
+
"lstrip": false,
|
40 |
+
"normalized": false,
|
41 |
+
"rstrip": true,
|
42 |
+
"single_word": false,
|
43 |
+
"special": true
|
44 |
+
},
|
45 |
+
"32002": {
|
46 |
+
"content": "<|placeholder1|>",
|
47 |
+
"lstrip": false,
|
48 |
+
"normalized": false,
|
49 |
+
"rstrip": true,
|
50 |
+
"single_word": false,
|
51 |
+
"special": true
|
52 |
+
},
|
53 |
+
"32003": {
|
54 |
+
"content": "<|placeholder2|>",
|
55 |
+
"lstrip": false,
|
56 |
+
"normalized": false,
|
57 |
+
"rstrip": true,
|
58 |
+
"single_word": false,
|
59 |
+
"special": true
|
60 |
+
},
|
61 |
+
"32004": {
|
62 |
+
"content": "<|placeholder3|>",
|
63 |
+
"lstrip": false,
|
64 |
+
"normalized": false,
|
65 |
+
"rstrip": true,
|
66 |
+
"single_word": false,
|
67 |
+
"special": true
|
68 |
+
},
|
69 |
+
"32005": {
|
70 |
+
"content": "<|placeholder4|>",
|
71 |
+
"lstrip": false,
|
72 |
+
"normalized": false,
|
73 |
+
"rstrip": true,
|
74 |
+
"single_word": false,
|
75 |
+
"special": true
|
76 |
+
},
|
77 |
+
"32006": {
|
78 |
+
"content": "<|system|>",
|
79 |
+
"lstrip": false,
|
80 |
+
"normalized": false,
|
81 |
+
"rstrip": true,
|
82 |
+
"single_word": false,
|
83 |
+
"special": true
|
84 |
+
},
|
85 |
+
"32007": {
|
86 |
+
"content": "<|end|>",
|
87 |
+
"lstrip": false,
|
88 |
+
"normalized": false,
|
89 |
+
"rstrip": true,
|
90 |
+
"single_word": false,
|
91 |
+
"special": true
|
92 |
+
},
|
93 |
+
"32008": {
|
94 |
+
"content": "<|placeholder5|>",
|
95 |
+
"lstrip": false,
|
96 |
+
"normalized": false,
|
97 |
+
"rstrip": true,
|
98 |
+
"single_word": false,
|
99 |
+
"special": true
|
100 |
+
},
|
101 |
+
"32009": {
|
102 |
+
"content": "<|placeholder6|>",
|
103 |
+
"lstrip": false,
|
104 |
+
"normalized": false,
|
105 |
+
"rstrip": true,
|
106 |
+
"single_word": false,
|
107 |
+
"special": true
|
108 |
+
},
|
109 |
+
"32010": {
|
110 |
+
"content": "<|user|>",
|
111 |
+
"lstrip": false,
|
112 |
+
"normalized": false,
|
113 |
+
"rstrip": true,
|
114 |
+
"single_word": false,
|
115 |
+
"special": true
|
116 |
+
}
|
117 |
+
},
|
118 |
+
"bos_token": "<s>",
|
119 |
+
"chat_template": "{{ bos_token }}{% for message in messages %}{% if (message['role'] == 'user') %}{{'<|user|>' + '\n' + message['content'] + '<|end|>' + '\n' + '<|assistant|>' + '\n'}}{% elif (message['role'] == 'assistant') %}{{message['content'] + '<|end|>' + '\n'}}{% endif %}{% endfor %}",
|
120 |
+
"clean_up_tokenization_spaces": false,
|
121 |
+
"eos_token": "<|endoftext|>",
|
122 |
+
"legacy": false,
|
123 |
+
"model_max_length": 2048,
|
124 |
+
"pad_token": "<|endoftext|>",
|
125 |
+
"padding_side": "left",
|
126 |
+
"sp_model_kwargs": {},
|
127 |
+
"tokenizer_class": "LlamaTokenizer",
|
128 |
+
"unk_token": "<unk>",
|
129 |
+
"use_default_system_prompt": false
|
130 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 0.9981298423724285,
|
3 |
+
"total_flos": 0.0,
|
4 |
+
"train_loss": 1.6566840494470372,
|
5 |
+
"train_runtime": 5417.7619,
|
6 |
+
"train_samples": 59876,
|
7 |
+
"train_samples_per_second": 11.052,
|
8 |
+
"train_steps_per_second": 0.086
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1468 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 0.9981298423724285,
|
5 |
+
"eval_steps": 400,
|
6 |
+
"global_step": 467,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.0021373230029388193,
|
13 |
+
"grad_norm": 42.099143078937416,
|
14 |
+
"learning_rate": 1.0638297872340425e-08,
|
15 |
+
"logits/chosen": 2.2472424507141113,
|
16 |
+
"logits/rejected": 2.264739513397217,
|
17 |
+
"logps/chosen": -0.8019136786460876,
|
18 |
+
"logps/rejected": -0.7812027335166931,
|
19 |
+
"loss": 1.7463,
|
20 |
+
"rewards/accuracies": 0.4375,
|
21 |
+
"rewards/chosen": -2.004784345626831,
|
22 |
+
"rewards/margins": -0.05177761986851692,
|
23 |
+
"rewards/rejected": -1.953006625175476,
|
24 |
+
"step": 1
|
25 |
+
},
|
26 |
+
{
|
27 |
+
"epoch": 0.010686615014694095,
|
28 |
+
"grad_norm": 43.49437439863032,
|
29 |
+
"learning_rate": 5.3191489361702123e-08,
|
30 |
+
"logits/chosen": 2.1162009239196777,
|
31 |
+
"logits/rejected": 2.0480892658233643,
|
32 |
+
"logps/chosen": -1.0361735820770264,
|
33 |
+
"logps/rejected": -1.1913045644760132,
|
34 |
+
"loss": 1.9184,
|
35 |
+
"rewards/accuracies": 0.453125,
|
36 |
+
"rewards/chosen": -2.5904340744018555,
|
37 |
+
"rewards/margins": 0.38782745599746704,
|
38 |
+
"rewards/rejected": -2.9782614707946777,
|
39 |
+
"step": 5
|
40 |
+
},
|
41 |
+
{
|
42 |
+
"epoch": 0.02137323002938819,
|
43 |
+
"grad_norm": 73.58193785863429,
|
44 |
+
"learning_rate": 1.0638297872340425e-07,
|
45 |
+
"logits/chosen": 2.0012879371643066,
|
46 |
+
"logits/rejected": 1.9172722101211548,
|
47 |
+
"logps/chosen": -1.5639755725860596,
|
48 |
+
"logps/rejected": -1.5803353786468506,
|
49 |
+
"loss": 1.9077,
|
50 |
+
"rewards/accuracies": 0.5249999761581421,
|
51 |
+
"rewards/chosen": -3.9099388122558594,
|
52 |
+
"rewards/margins": 0.04089919477701187,
|
53 |
+
"rewards/rejected": -3.950838088989258,
|
54 |
+
"step": 10
|
55 |
+
},
|
56 |
+
{
|
57 |
+
"epoch": 0.03205984504408229,
|
58 |
+
"grad_norm": 18.887588182169267,
|
59 |
+
"learning_rate": 1.5957446808510638e-07,
|
60 |
+
"logits/chosen": 1.9121309518814087,
|
61 |
+
"logits/rejected": 1.8687114715576172,
|
62 |
+
"logps/chosen": -1.1899975538253784,
|
63 |
+
"logps/rejected": -1.2847260236740112,
|
64 |
+
"loss": 1.832,
|
65 |
+
"rewards/accuracies": 0.550000011920929,
|
66 |
+
"rewards/chosen": -2.974993944168091,
|
67 |
+
"rewards/margins": 0.2368215024471283,
|
68 |
+
"rewards/rejected": -3.211815357208252,
|
69 |
+
"step": 15
|
70 |
+
},
|
71 |
+
{
|
72 |
+
"epoch": 0.04274646005877638,
|
73 |
+
"grad_norm": 8.819209234657059,
|
74 |
+
"learning_rate": 2.127659574468085e-07,
|
75 |
+
"logits/chosen": 1.8210818767547607,
|
76 |
+
"logits/rejected": 1.8469772338867188,
|
77 |
+
"logps/chosen": -1.21903657913208,
|
78 |
+
"logps/rejected": -1.7520463466644287,
|
79 |
+
"loss": 1.7601,
|
80 |
+
"rewards/accuracies": 0.7124999761581421,
|
81 |
+
"rewards/chosen": -3.0475919246673584,
|
82 |
+
"rewards/margins": 1.3325239419937134,
|
83 |
+
"rewards/rejected": -4.380115509033203,
|
84 |
+
"step": 20
|
85 |
+
},
|
86 |
+
{
|
87 |
+
"epoch": 0.053433075073470476,
|
88 |
+
"grad_norm": 33.18790384628351,
|
89 |
+
"learning_rate": 2.659574468085106e-07,
|
90 |
+
"logits/chosen": 2.0562968254089355,
|
91 |
+
"logits/rejected": 2.0691914558410645,
|
92 |
+
"logps/chosen": -1.0988093614578247,
|
93 |
+
"logps/rejected": -1.1313331127166748,
|
94 |
+
"loss": 1.815,
|
95 |
+
"rewards/accuracies": 0.5249999761581421,
|
96 |
+
"rewards/chosen": -2.747023344039917,
|
97 |
+
"rewards/margins": 0.08130944520235062,
|
98 |
+
"rewards/rejected": -2.8283333778381348,
|
99 |
+
"step": 25
|
100 |
+
},
|
101 |
+
{
|
102 |
+
"epoch": 0.06411969008816458,
|
103 |
+
"grad_norm": 35.16962019463958,
|
104 |
+
"learning_rate": 3.1914893617021275e-07,
|
105 |
+
"logits/chosen": 1.9061098098754883,
|
106 |
+
"logits/rejected": 1.9218835830688477,
|
107 |
+
"logps/chosen": -1.2370525598526,
|
108 |
+
"logps/rejected": -1.4245789051055908,
|
109 |
+
"loss": 1.6802,
|
110 |
+
"rewards/accuracies": 0.4749999940395355,
|
111 |
+
"rewards/chosen": -3.0926311016082764,
|
112 |
+
"rewards/margins": 0.46881628036499023,
|
113 |
+
"rewards/rejected": -3.5614471435546875,
|
114 |
+
"step": 30
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.07480630510285867,
|
118 |
+
"grad_norm": 44.11653715682163,
|
119 |
+
"learning_rate": 3.7234042553191484e-07,
|
120 |
+
"logits/chosen": 1.9321657419204712,
|
121 |
+
"logits/rejected": 1.955640435218811,
|
122 |
+
"logps/chosen": -1.344245433807373,
|
123 |
+
"logps/rejected": -1.36150062084198,
|
124 |
+
"loss": 1.8113,
|
125 |
+
"rewards/accuracies": 0.42500001192092896,
|
126 |
+
"rewards/chosen": -3.3606135845184326,
|
127 |
+
"rewards/margins": 0.0431380569934845,
|
128 |
+
"rewards/rejected": -3.4037513732910156,
|
129 |
+
"step": 35
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 0.08549292011755276,
|
133 |
+
"grad_norm": 36.76549905569564,
|
134 |
+
"learning_rate": 4.25531914893617e-07,
|
135 |
+
"logits/chosen": 1.8915932178497314,
|
136 |
+
"logits/rejected": 1.9206931591033936,
|
137 |
+
"logps/chosen": -1.2968499660491943,
|
138 |
+
"logps/rejected": -1.3007665872573853,
|
139 |
+
"loss": 1.8425,
|
140 |
+
"rewards/accuracies": 0.5874999761581421,
|
141 |
+
"rewards/chosen": -3.2421250343322754,
|
142 |
+
"rewards/margins": 0.009791326709091663,
|
143 |
+
"rewards/rejected": -3.2519161701202393,
|
144 |
+
"step": 40
|
145 |
+
},
|
146 |
+
{
|
147 |
+
"epoch": 0.09617953513224686,
|
148 |
+
"grad_norm": 17.026888436798433,
|
149 |
+
"learning_rate": 4.787234042553192e-07,
|
150 |
+
"logits/chosen": 1.8335082530975342,
|
151 |
+
"logits/rejected": 1.803972601890564,
|
152 |
+
"logps/chosen": -1.1560873985290527,
|
153 |
+
"logps/rejected": -1.391876459121704,
|
154 |
+
"loss": 1.748,
|
155 |
+
"rewards/accuracies": 0.550000011920929,
|
156 |
+
"rewards/chosen": -2.8902182579040527,
|
157 |
+
"rewards/margins": 0.5894726514816284,
|
158 |
+
"rewards/rejected": -3.4796910285949707,
|
159 |
+
"step": 45
|
160 |
+
},
|
161 |
+
{
|
162 |
+
"epoch": 0.10686615014694095,
|
163 |
+
"grad_norm": 12.410804374401861,
|
164 |
+
"learning_rate": 4.999370587356267e-07,
|
165 |
+
"logits/chosen": 1.864829659461975,
|
166 |
+
"logits/rejected": 1.8258002996444702,
|
167 |
+
"logps/chosen": -1.0499851703643799,
|
168 |
+
"logps/rejected": -1.454998254776001,
|
169 |
+
"loss": 1.5838,
|
170 |
+
"rewards/accuracies": 0.5249999761581421,
|
171 |
+
"rewards/chosen": -2.6249632835388184,
|
172 |
+
"rewards/margins": 1.0125324726104736,
|
173 |
+
"rewards/rejected": -3.637495517730713,
|
174 |
+
"step": 50
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 0.11755276516163506,
|
178 |
+
"grad_norm": 16.97890172462393,
|
179 |
+
"learning_rate": 4.995525324419337e-07,
|
180 |
+
"logits/chosen": 1.8395353555679321,
|
181 |
+
"logits/rejected": 1.8361022472381592,
|
182 |
+
"logps/chosen": -0.9045565724372864,
|
183 |
+
"logps/rejected": -1.1536071300506592,
|
184 |
+
"loss": 1.6796,
|
185 |
+
"rewards/accuracies": 0.550000011920929,
|
186 |
+
"rewards/chosen": -2.2613916397094727,
|
187 |
+
"rewards/margins": 0.6226258873939514,
|
188 |
+
"rewards/rejected": -2.8840174674987793,
|
189 |
+
"step": 55
|
190 |
+
},
|
191 |
+
{
|
192 |
+
"epoch": 0.12823938017632916,
|
193 |
+
"grad_norm": 18.13814698710649,
|
194 |
+
"learning_rate": 4.988189843662815e-07,
|
195 |
+
"logits/chosen": 1.9378535747528076,
|
196 |
+
"logits/rejected": 1.9659252166748047,
|
197 |
+
"logps/chosen": -1.1384503841400146,
|
198 |
+
"logps/rejected": -1.0124545097351074,
|
199 |
+
"loss": 1.7608,
|
200 |
+
"rewards/accuracies": 0.550000011920929,
|
201 |
+
"rewards/chosen": -2.846126079559326,
|
202 |
+
"rewards/margins": -0.3149900436401367,
|
203 |
+
"rewards/rejected": -2.5311360359191895,
|
204 |
+
"step": 60
|
205 |
+
},
|
206 |
+
{
|
207 |
+
"epoch": 0.13892599519102325,
|
208 |
+
"grad_norm": 31.166552691677705,
|
209 |
+
"learning_rate": 4.977374404419837e-07,
|
210 |
+
"logits/chosen": 2.1393017768859863,
|
211 |
+
"logits/rejected": 2.103208303451538,
|
212 |
+
"logps/chosen": -0.8744574785232544,
|
213 |
+
"logps/rejected": -0.9523324966430664,
|
214 |
+
"loss": 1.6552,
|
215 |
+
"rewards/accuracies": 0.612500011920929,
|
216 |
+
"rewards/chosen": -2.186143636703491,
|
217 |
+
"rewards/margins": 0.19468756020069122,
|
218 |
+
"rewards/rejected": -2.380831241607666,
|
219 |
+
"step": 65
|
220 |
+
},
|
221 |
+
{
|
222 |
+
"epoch": 0.14961261020571734,
|
223 |
+
"grad_norm": 52.728695245981314,
|
224 |
+
"learning_rate": 4.963094133060148e-07,
|
225 |
+
"logits/chosen": 2.0067527294158936,
|
226 |
+
"logits/rejected": 1.9597012996673584,
|
227 |
+
"logps/chosen": -0.9174124598503113,
|
228 |
+
"logps/rejected": -1.1131672859191895,
|
229 |
+
"loss": 1.7212,
|
230 |
+
"rewards/accuracies": 0.550000011920929,
|
231 |
+
"rewards/chosen": -2.2935314178466797,
|
232 |
+
"rewards/margins": 0.4893869459629059,
|
233 |
+
"rewards/rejected": -2.7829184532165527,
|
234 |
+
"step": 70
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 0.16029922522041143,
|
238 |
+
"grad_norm": 46.624386507512185,
|
239 |
+
"learning_rate": 4.945369001834514e-07,
|
240 |
+
"logits/chosen": 1.8767681121826172,
|
241 |
+
"logits/rejected": 1.8939720392227173,
|
242 |
+
"logps/chosen": -0.9977117776870728,
|
243 |
+
"logps/rejected": -1.1487284898757935,
|
244 |
+
"loss": 1.6147,
|
245 |
+
"rewards/accuracies": 0.5874999761581421,
|
246 |
+
"rewards/chosen": -2.494279384613037,
|
247 |
+
"rewards/margins": 0.3775419294834137,
|
248 |
+
"rewards/rejected": -2.8718209266662598,
|
249 |
+
"step": 75
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 0.17098584023510552,
|
253 |
+
"grad_norm": 13.45137638802328,
|
254 |
+
"learning_rate": 4.924223800941717e-07,
|
255 |
+
"logits/chosen": 1.912676453590393,
|
256 |
+
"logits/rejected": 1.8263626098632812,
|
257 |
+
"logps/chosen": -0.9859241247177124,
|
258 |
+
"logps/rejected": -1.0560874938964844,
|
259 |
+
"loss": 1.651,
|
260 |
+
"rewards/accuracies": 0.5375000238418579,
|
261 |
+
"rewards/chosen": -2.464810609817505,
|
262 |
+
"rewards/margins": 0.17540840804576874,
|
263 |
+
"rewards/rejected": -2.640218734741211,
|
264 |
+
"step": 80
|
265 |
+
},
|
266 |
+
{
|
267 |
+
"epoch": 0.18167245524979964,
|
268 |
+
"grad_norm": 23.876540575284714,
|
269 |
+
"learning_rate": 4.899688103857222e-07,
|
270 |
+
"logits/chosen": 1.9252907037734985,
|
271 |
+
"logits/rejected": 1.9396631717681885,
|
272 |
+
"logps/chosen": -0.9312021136283875,
|
273 |
+
"logps/rejected": -1.1102412939071655,
|
274 |
+
"loss": 1.618,
|
275 |
+
"rewards/accuracies": 0.550000011920929,
|
276 |
+
"rewards/chosen": -2.328005313873291,
|
277 |
+
"rewards/margins": 0.44759804010391235,
|
278 |
+
"rewards/rejected": -2.7756032943725586,
|
279 |
+
"step": 85
|
280 |
+
},
|
281 |
+
{
|
282 |
+
"epoch": 0.19235907026449373,
|
283 |
+
"grad_norm": 7.705041238913054,
|
284 |
+
"learning_rate": 4.871796225971999e-07,
|
285 |
+
"logits/chosen": 1.9238191843032837,
|
286 |
+
"logits/rejected": 1.8918044567108154,
|
287 |
+
"logps/chosen": -0.9711843729019165,
|
288 |
+
"logps/rejected": -1.096145749092102,
|
289 |
+
"loss": 1.6624,
|
290 |
+
"rewards/accuracies": 0.512499988079071,
|
291 |
+
"rewards/chosen": -2.4279608726501465,
|
292 |
+
"rewards/margins": 0.3124033510684967,
|
293 |
+
"rewards/rejected": -2.7403643131256104,
|
294 |
+
"step": 90
|
295 |
+
},
|
296 |
+
{
|
297 |
+
"epoch": 0.20304568527918782,
|
298 |
+
"grad_norm": 16.98405162222745,
|
299 |
+
"learning_rate": 4.840587176599343e-07,
|
300 |
+
"logits/chosen": 2.0161519050598145,
|
301 |
+
"logits/rejected": 2.037721872329712,
|
302 |
+
"logps/chosen": -0.9618253707885742,
|
303 |
+
"logps/rejected": -1.1678836345672607,
|
304 |
+
"loss": 1.6553,
|
305 |
+
"rewards/accuracies": 0.637499988079071,
|
306 |
+
"rewards/chosen": -2.4045631885528564,
|
307 |
+
"rewards/margins": 0.5151457786560059,
|
308 |
+
"rewards/rejected": -2.9197089672088623,
|
309 |
+
"step": 95
|
310 |
+
},
|
311 |
+
{
|
312 |
+
"epoch": 0.2137323002938819,
|
313 |
+
"grad_norm": 26.59166513674174,
|
314 |
+
"learning_rate": 4.806104604416823e-07,
|
315 |
+
"logits/chosen": 1.9551931619644165,
|
316 |
+
"logits/rejected": 1.9663063287734985,
|
317 |
+
"logps/chosen": -0.9686329960823059,
|
318 |
+
"logps/rejected": -0.964672863483429,
|
319 |
+
"loss": 1.6952,
|
320 |
+
"rewards/accuracies": 0.48750001192092896,
|
321 |
+
"rewards/chosen": -2.4215824604034424,
|
322 |
+
"rewards/margins": -0.009900185279548168,
|
323 |
+
"rewards/rejected": -2.41168212890625,
|
324 |
+
"step": 100
|
325 |
+
},
|
326 |
+
{
|
327 |
+
"epoch": 0.224418915308576,
|
328 |
+
"grad_norm": 22.597015867331866,
|
329 |
+
"learning_rate": 4.768396736419662e-07,
|
330 |
+
"logits/chosen": 2.0987541675567627,
|
331 |
+
"logits/rejected": 2.1704182624816895,
|
332 |
+
"logps/chosen": -0.8102301359176636,
|
333 |
+
"logps/rejected": -1.0246448516845703,
|
334 |
+
"loss": 1.6353,
|
335 |
+
"rewards/accuracies": 0.574999988079071,
|
336 |
+
"rewards/chosen": -2.0255753993988037,
|
337 |
+
"rewards/margins": 0.5360367894172668,
|
338 |
+
"rewards/rejected": -2.561612367630005,
|
339 |
+
"step": 105
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 0.2351055303232701,
|
343 |
+
"grad_norm": 7.710202530561847,
|
344 |
+
"learning_rate": 4.7275163104709194e-07,
|
345 |
+
"logits/chosen": 1.8211323022842407,
|
346 |
+
"logits/rejected": 1.8467600345611572,
|
347 |
+
"logps/chosen": -0.8923318982124329,
|
348 |
+
"logps/rejected": -0.9886472821235657,
|
349 |
+
"loss": 1.6826,
|
350 |
+
"rewards/accuracies": 0.612500011920929,
|
351 |
+
"rewards/chosen": -2.2308297157287598,
|
352 |
+
"rewards/margins": 0.24078825116157532,
|
353 |
+
"rewards/rejected": -2.4716179370880127,
|
354 |
+
"step": 110
|
355 |
+
},
|
356 |
+
{
|
357 |
+
"epoch": 0.2457921453379642,
|
358 |
+
"grad_norm": 9.211974257959199,
|
359 |
+
"learning_rate": 4.683520501542824e-07,
|
360 |
+
"logits/chosen": 1.822178840637207,
|
361 |
+
"logits/rejected": 1.8460540771484375,
|
362 |
+
"logps/chosen": -0.937783420085907,
|
363 |
+
"logps/rejected": -0.9862428903579712,
|
364 |
+
"loss": 1.6087,
|
365 |
+
"rewards/accuracies": 0.48750001192092896,
|
366 |
+
"rewards/chosen": -2.34445858001709,
|
367 |
+
"rewards/margins": 0.12114912271499634,
|
368 |
+
"rewards/rejected": -2.4656076431274414,
|
369 |
+
"step": 115
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 0.2564787603526583,
|
373 |
+
"grad_norm": 17.97332526477704,
|
374 |
+
"learning_rate": 4.636470841752404e-07,
|
375 |
+
"logits/chosen": 1.9925601482391357,
|
376 |
+
"logits/rejected": 2.0132079124450684,
|
377 |
+
"logps/chosen": -0.8938210606575012,
|
378 |
+
"logps/rejected": -0.9105901718139648,
|
379 |
+
"loss": 1.6647,
|
380 |
+
"rewards/accuracies": 0.4749999940395355,
|
381 |
+
"rewards/chosen": -2.2345528602600098,
|
382 |
+
"rewards/margins": 0.04192250221967697,
|
383 |
+
"rewards/rejected": -2.276475429534912,
|
384 |
+
"step": 120
|
385 |
+
},
|
386 |
+
{
|
387 |
+
"epoch": 0.2671653753673524,
|
388 |
+
"grad_norm": 21.241131663632817,
|
389 |
+
"learning_rate": 4.5864331343032565e-07,
|
390 |
+
"logits/chosen": 1.821298360824585,
|
391 |
+
"logits/rejected": 1.8526639938354492,
|
392 |
+
"logps/chosen": -0.9126847386360168,
|
393 |
+
"logps/rejected": -1.0501748323440552,
|
394 |
+
"loss": 1.7558,
|
395 |
+
"rewards/accuracies": 0.625,
|
396 |
+
"rewards/chosen": -2.281712055206299,
|
397 |
+
"rewards/margins": 0.3437250852584839,
|
398 |
+
"rewards/rejected": -2.6254372596740723,
|
399 |
+
"step": 125
|
400 |
+
},
|
401 |
+
{
|
402 |
+
"epoch": 0.2778519903820465,
|
403 |
+
"grad_norm": 18.947973877308787,
|
404 |
+
"learning_rate": 4.533477361453819e-07,
|
405 |
+
"logits/chosen": 2.178891658782959,
|
406 |
+
"logits/rejected": 2.143307685852051,
|
407 |
+
"logps/chosen": -0.9617658853530884,
|
408 |
+
"logps/rejected": -1.223284125328064,
|
409 |
+
"loss": 1.6892,
|
410 |
+
"rewards/accuracies": 0.5249999761581421,
|
411 |
+
"rewards/chosen": -2.404414653778076,
|
412 |
+
"rewards/margins": 0.6537958383560181,
|
413 |
+
"rewards/rejected": -3.0582103729248047,
|
414 |
+
"step": 130
|
415 |
+
},
|
416 |
+
{
|
417 |
+
"epoch": 0.2885386053967406,
|
418 |
+
"grad_norm": 13.76012227966358,
|
419 |
+
"learning_rate": 4.4776775866408533e-07,
|
420 |
+
"logits/chosen": 1.8531124591827393,
|
421 |
+
"logits/rejected": 1.8338654041290283,
|
422 |
+
"logps/chosen": -0.8667293787002563,
|
423 |
+
"logps/rejected": -0.9339718818664551,
|
424 |
+
"loss": 1.658,
|
425 |
+
"rewards/accuracies": 0.5249999761581421,
|
426 |
+
"rewards/chosen": -2.166823148727417,
|
427 |
+
"rewards/margins": 0.16810639202594757,
|
428 |
+
"rewards/rejected": -2.334929943084717,
|
429 |
+
"step": 135
|
430 |
+
},
|
431 |
+
{
|
432 |
+
"epoch": 0.2992252204114347,
|
433 |
+
"grad_norm": 12.832812020829884,
|
434 |
+
"learning_rate": 4.4191118508950277e-07,
|
435 |
+
"logits/chosen": 1.8038040399551392,
|
436 |
+
"logits/rejected": 1.7551212310791016,
|
437 |
+
"logps/chosen": -0.8944159746170044,
|
438 |
+
"logps/rejected": -1.0033025741577148,
|
439 |
+
"loss": 1.6771,
|
440 |
+
"rewards/accuracies": 0.637499988079071,
|
441 |
+
"rewards/chosen": -2.236039638519287,
|
442 |
+
"rewards/margins": 0.2722164988517761,
|
443 |
+
"rewards/rejected": -2.508256435394287,
|
444 |
+
"step": 140
|
445 |
+
},
|
446 |
+
{
|
447 |
+
"epoch": 0.30991183542612877,
|
448 |
+
"grad_norm": 9.09964498419542,
|
449 |
+
"learning_rate": 4.357862063693485e-07,
|
450 |
+
"logits/chosen": 2.027944564819336,
|
451 |
+
"logits/rejected": 2.005831718444824,
|
452 |
+
"logps/chosen": -0.8879717588424683,
|
453 |
+
"logps/rejected": -0.9542590975761414,
|
454 |
+
"loss": 1.7133,
|
455 |
+
"rewards/accuracies": 0.550000011920929,
|
456 |
+
"rewards/chosen": -2.2199294567108154,
|
457 |
+
"rewards/margins": 0.16571831703186035,
|
458 |
+
"rewards/rejected": -2.3856475353240967,
|
459 |
+
"step": 145
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 0.32059845044082286,
|
463 |
+
"grad_norm": 24.685014270210495,
|
464 |
+
"learning_rate": 4.294013888402029e-07,
|
465 |
+
"logits/chosen": 2.138366222381592,
|
466 |
+
"logits/rejected": 2.2007527351379395,
|
467 |
+
"logps/chosen": -0.8750768899917603,
|
468 |
+
"logps/rejected": -1.0953787565231323,
|
469 |
+
"loss": 1.6625,
|
470 |
+
"rewards/accuracies": 0.574999988079071,
|
471 |
+
"rewards/chosen": -2.187692165374756,
|
472 |
+
"rewards/margins": 0.5507549047470093,
|
473 |
+
"rewards/rejected": -2.7384471893310547,
|
474 |
+
"step": 150
|
475 |
+
},
|
476 |
+
{
|
477 |
+
"epoch": 0.33128506545551695,
|
478 |
+
"grad_norm": 24.61662782262514,
|
479 |
+
"learning_rate": 4.227656622467162e-07,
|
480 |
+
"logits/chosen": 1.911320686340332,
|
481 |
+
"logits/rejected": 1.9191503524780273,
|
482 |
+
"logps/chosen": -0.8599516153335571,
|
483 |
+
"logps/rejected": -1.1553843021392822,
|
484 |
+
"loss": 1.6495,
|
485 |
+
"rewards/accuracies": 0.550000011920929,
|
486 |
+
"rewards/chosen": -2.149878978729248,
|
487 |
+
"rewards/margins": 0.7385817766189575,
|
488 |
+
"rewards/rejected": -2.888460874557495,
|
489 |
+
"step": 155
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"epoch": 0.34197168047021104,
|
493 |
+
"grad_norm": 9.846236022517108,
|
494 |
+
"learning_rate": 4.158883072525528e-07,
|
495 |
+
"logits/chosen": 2.061549425125122,
|
496 |
+
"logits/rejected": 2.0822834968566895,
|
497 |
+
"logps/chosen": -0.8990328907966614,
|
498 |
+
"logps/rejected": -0.985998809337616,
|
499 |
+
"loss": 1.6348,
|
500 |
+
"rewards/accuracies": 0.4625000059604645,
|
501 |
+
"rewards/chosen": -2.247582197189331,
|
502 |
+
"rewards/margins": 0.21741452813148499,
|
503 |
+
"rewards/rejected": -2.464996814727783,
|
504 |
+
"step": 160
|
505 |
+
},
|
506 |
+
{
|
507 |
+
"epoch": 0.3526582954849052,
|
508 |
+
"grad_norm": 19.106656482086205,
|
509 |
+
"learning_rate": 4.087789424605447e-07,
|
510 |
+
"logits/chosen": 2.265204906463623,
|
511 |
+
"logits/rejected": 2.245617151260376,
|
512 |
+
"logps/chosen": -0.818160355091095,
|
513 |
+
"logps/rejected": -0.9623705744743347,
|
514 |
+
"loss": 1.618,
|
515 |
+
"rewards/accuracies": 0.5,
|
516 |
+
"rewards/chosen": -2.045400857925415,
|
517 |
+
"rewards/margins": 0.3605257570743561,
|
518 |
+
"rewards/rejected": -2.405926465988159,
|
519 |
+
"step": 165
|
520 |
+
},
|
521 |
+
{
|
522 |
+
"epoch": 0.36334491049959927,
|
523 |
+
"grad_norm": 14.919110153191296,
|
524 |
+
"learning_rate": 4.0144751096020497e-07,
|
525 |
+
"logits/chosen": 1.9928451776504517,
|
526 |
+
"logits/rejected": 1.9697974920272827,
|
527 |
+
"logps/chosen": -1.013695478439331,
|
528 |
+
"logps/rejected": -1.2301123142242432,
|
529 |
+
"loss": 1.5848,
|
530 |
+
"rewards/accuracies": 0.550000011920929,
|
531 |
+
"rewards/chosen": -2.534238815307617,
|
532 |
+
"rewards/margins": 0.5410416722297668,
|
533 |
+
"rewards/rejected": -3.0752806663513184,
|
534 |
+
"step": 170
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 0.37403152551429336,
|
538 |
+
"grad_norm": 11.830869809507767,
|
539 |
+
"learning_rate": 3.939042664214184e-07,
|
540 |
+
"logits/chosen": 2.031292676925659,
|
541 |
+
"logits/rejected": 2.066436767578125,
|
542 |
+
"logps/chosen": -0.97381991147995,
|
543 |
+
"logps/rejected": -1.2112761735916138,
|
544 |
+
"loss": 1.5954,
|
545 |
+
"rewards/accuracies": 0.6000000238418579,
|
546 |
+
"rewards/chosen": -2.4345498085021973,
|
547 |
+
"rewards/margins": 0.5936408638954163,
|
548 |
+
"rewards/rejected": -3.0281906127929688,
|
549 |
+
"step": 175
|
550 |
+
},
|
551 |
+
{
|
552 |
+
"epoch": 0.38471814052898745,
|
553 |
+
"grad_norm": 14.628170167780459,
|
554 |
+
"learning_rate": 3.8615975875375676e-07,
|
555 |
+
"logits/chosen": 2.258387804031372,
|
556 |
+
"logits/rejected": 2.3218982219696045,
|
557 |
+
"logps/chosen": -1.0137016773223877,
|
558 |
+
"logps/rejected": -1.1529831886291504,
|
559 |
+
"loss": 1.6695,
|
560 |
+
"rewards/accuracies": 0.6499999761581421,
|
561 |
+
"rewards/chosen": -2.534254312515259,
|
562 |
+
"rewards/margins": 0.3482036888599396,
|
563 |
+
"rewards/rejected": -2.882457971572876,
|
564 |
+
"step": 180
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 0.39540475554368154,
|
568 |
+
"grad_norm": 16.49947715615614,
|
569 |
+
"learning_rate": 3.7822481935147655e-07,
|
570 |
+
"logits/chosen": 2.0560050010681152,
|
571 |
+
"logits/rejected": 2.1167311668395996,
|
572 |
+
"logps/chosen": -0.9939271211624146,
|
573 |
+
"logps/rejected": -1.1564804315567017,
|
574 |
+
"loss": 1.6754,
|
575 |
+
"rewards/accuracies": 0.5249999761581421,
|
576 |
+
"rewards/chosen": -2.4848177433013916,
|
577 |
+
"rewards/margins": 0.40638333559036255,
|
578 |
+
"rewards/rejected": -2.8912010192871094,
|
579 |
+
"step": 185
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 0.40609137055837563,
|
583 |
+
"grad_norm": 13.5081717360946,
|
584 |
+
"learning_rate": 3.7011054594483443e-07,
|
585 |
+
"logits/chosen": 2.1290013790130615,
|
586 |
+
"logits/rejected": 2.223310947418213,
|
587 |
+
"logps/chosen": -0.9123891592025757,
|
588 |
+
"logps/rejected": -1.1287665367126465,
|
589 |
+
"loss": 1.5831,
|
590 |
+
"rewards/accuracies": 0.675000011920929,
|
591 |
+
"rewards/chosen": -2.280972957611084,
|
592 |
+
"rewards/margins": 0.5409437417984009,
|
593 |
+
"rewards/rejected": -2.8219165802001953,
|
594 |
+
"step": 190
|
595 |
+
},
|
596 |
+
{
|
597 |
+
"epoch": 0.4167779855730697,
|
598 |
+
"grad_norm": 17.61863203205812,
|
599 |
+
"learning_rate": 3.618282870789081e-07,
|
600 |
+
"logits/chosen": 2.1478934288024902,
|
601 |
+
"logits/rejected": 2.1439719200134277,
|
602 |
+
"logps/chosen": -0.974346935749054,
|
603 |
+
"logps/rejected": -1.1090627908706665,
|
604 |
+
"loss": 1.64,
|
605 |
+
"rewards/accuracies": 0.5249999761581421,
|
606 |
+
"rewards/chosen": -2.4358673095703125,
|
607 |
+
"rewards/margins": 0.33678966760635376,
|
608 |
+
"rewards/rejected": -2.7726569175720215,
|
609 |
+
"step": 195
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"epoch": 0.4274646005877638,
|
613 |
+
"grad_norm": 12.457201720784246,
|
614 |
+
"learning_rate": 3.5338962624163016e-07,
|
615 |
+
"logits/chosen": 2.2304863929748535,
|
616 |
+
"logits/rejected": 2.2417056560516357,
|
617 |
+
"logps/chosen": -0.9161917567253113,
|
618 |
+
"logps/rejected": -1.0349229574203491,
|
619 |
+
"loss": 1.672,
|
620 |
+
"rewards/accuracies": 0.6499999761581421,
|
621 |
+
"rewards/chosen": -2.2904791831970215,
|
622 |
+
"rewards/margins": 0.29682818055152893,
|
623 |
+
"rewards/rejected": -2.5873074531555176,
|
624 |
+
"step": 200
|
625 |
+
},
|
626 |
+
{
|
627 |
+
"epoch": 0.4381512156024579,
|
628 |
+
"grad_norm": 25.419906192801392,
|
629 |
+
"learning_rate": 3.448063656632321e-07,
|
630 |
+
"logits/chosen": 2.2168259620666504,
|
631 |
+
"logits/rejected": 2.2494800090789795,
|
632 |
+
"logps/chosen": -0.9495501518249512,
|
633 |
+
"logps/rejected": -1.0588171482086182,
|
634 |
+
"loss": 1.6462,
|
635 |
+
"rewards/accuracies": 0.550000011920929,
|
636 |
+
"rewards/chosen": -2.373875379562378,
|
637 |
+
"rewards/margins": 0.2731671929359436,
|
638 |
+
"rewards/rejected": -2.647042751312256,
|
639 |
+
"step": 205
|
640 |
+
},
|
641 |
+
{
|
642 |
+
"epoch": 0.448837830617152,
|
643 |
+
"grad_norm": 31.43495532901771,
|
644 |
+
"learning_rate": 3.360905098097587e-07,
|
645 |
+
"logits/chosen": 2.0355324745178223,
|
646 |
+
"logits/rejected": 2.0058634281158447,
|
647 |
+
"logps/chosen": -0.9030116200447083,
|
648 |
+
"logps/rejected": -1.100672960281372,
|
649 |
+
"loss": 1.7332,
|
650 |
+
"rewards/accuracies": 0.5249999761581421,
|
651 |
+
"rewards/chosen": -2.2575290203094482,
|
652 |
+
"rewards/margins": 0.49415332078933716,
|
653 |
+
"rewards/rejected": -2.751682758331299,
|
654 |
+
"step": 210
|
655 |
+
},
|
656 |
+
{
|
657 |
+
"epoch": 0.45952444563184613,
|
658 |
+
"grad_norm": 9.866165199153304,
|
659 |
+
"learning_rate": 3.272542485937368e-07,
|
660 |
+
"logits/chosen": 2.188046932220459,
|
661 |
+
"logits/rejected": 2.209195375442505,
|
662 |
+
"logps/chosen": -0.8452509045600891,
|
663 |
+
"logps/rejected": -0.9856406450271606,
|
664 |
+
"loss": 1.6173,
|
665 |
+
"rewards/accuracies": 0.574999988079071,
|
666 |
+
"rewards/chosen": -2.1131272315979004,
|
667 |
+
"rewards/margins": 0.3509744107723236,
|
668 |
+
"rewards/rejected": -2.464101791381836,
|
669 |
+
"step": 215
|
670 |
+
},
|
671 |
+
{
|
672 |
+
"epoch": 0.4702110606465402,
|
673 |
+
"grad_norm": 11.268271977292386,
|
674 |
+
"learning_rate": 3.1830994032548e-07,
|
675 |
+
"logits/chosen": 2.2558753490448,
|
676 |
+
"logits/rejected": 2.271909475326538,
|
677 |
+
"logps/chosen": -0.9621235728263855,
|
678 |
+
"logps/rejected": -1.0470136404037476,
|
679 |
+
"loss": 1.6623,
|
680 |
+
"rewards/accuracies": 0.5625,
|
681 |
+
"rewards/chosen": -2.405308961868286,
|
682 |
+
"rewards/margins": 0.21222510933876038,
|
683 |
+
"rewards/rejected": -2.6175341606140137,
|
684 |
+
"step": 220
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"epoch": 0.4808976756612343,
|
688 |
+
"grad_norm": 8.220154130520008,
|
689 |
+
"learning_rate": 3.0927009442887437e-07,
|
690 |
+
"logits/chosen": 2.033316135406494,
|
691 |
+
"logits/rejected": 2.0451745986938477,
|
692 |
+
"logps/chosen": -0.9348015785217285,
|
693 |
+
"logps/rejected": -0.9307538866996765,
|
694 |
+
"loss": 1.7025,
|
695 |
+
"rewards/accuracies": 0.512499988079071,
|
696 |
+
"rewards/chosen": -2.3370039463043213,
|
697 |
+
"rewards/margins": -0.010119226761162281,
|
698 |
+
"rewards/rejected": -2.3268847465515137,
|
699 |
+
"step": 225
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 0.4915842906759284,
|
703 |
+
"grad_norm": 8.761484125412297,
|
704 |
+
"learning_rate": 3.001473539458182e-07,
|
705 |
+
"logits/chosen": 2.039243221282959,
|
706 |
+
"logits/rejected": 2.0847630500793457,
|
707 |
+
"logps/chosen": -0.9077224731445312,
|
708 |
+
"logps/rejected": -0.9697052240371704,
|
709 |
+
"loss": 1.6369,
|
710 |
+
"rewards/accuracies": 0.5375000238418579,
|
711 |
+
"rewards/chosen": -2.269306182861328,
|
712 |
+
"rewards/margins": 0.15495693683624268,
|
713 |
+
"rewards/rejected": -2.4242632389068604,
|
714 |
+
"step": 230
|
715 |
+
},
|
716 |
+
{
|
717 |
+
"epoch": 0.5022709056906225,
|
718 |
+
"grad_norm": 21.069821104434805,
|
719 |
+
"learning_rate": 2.909544778537844e-07,
|
720 |
+
"logits/chosen": 2.1515793800354004,
|
721 |
+
"logits/rejected": 2.0550239086151123,
|
722 |
+
"logps/chosen": -0.9554904699325562,
|
723 |
+
"logps/rejected": -1.0575495958328247,
|
724 |
+
"loss": 1.6471,
|
725 |
+
"rewards/accuracies": 0.4625000059604645,
|
726 |
+
"rewards/chosen": -2.3887264728546143,
|
727 |
+
"rewards/margins": 0.2551475167274475,
|
728 |
+
"rewards/rejected": -2.643873691558838,
|
729 |
+
"step": 235
|
730 |
+
},
|
731 |
+
{
|
732 |
+
"epoch": 0.5129575207053166,
|
733 |
+
"grad_norm": 18.45689011337618,
|
734 |
+
"learning_rate": 2.817043232212371e-07,
|
735 |
+
"logits/chosen": 2.3277316093444824,
|
736 |
+
"logits/rejected": 2.359060287475586,
|
737 |
+
"logps/chosen": -0.9175651669502258,
|
738 |
+
"logps/rejected": -1.075629472732544,
|
739 |
+
"loss": 1.5723,
|
740 |
+
"rewards/accuracies": 0.5249999761581421,
|
741 |
+
"rewards/chosen": -2.2939131259918213,
|
742 |
+
"rewards/margins": 0.39516064524650574,
|
743 |
+
"rewards/rejected": -2.6890735626220703,
|
744 |
+
"step": 240
|
745 |
+
},
|
746 |
+
{
|
747 |
+
"epoch": 0.5236441357200107,
|
748 |
+
"grad_norm": 5.945491870174955,
|
749 |
+
"learning_rate": 2.7240982722585837e-07,
|
750 |
+
"logits/chosen": 2.2272980213165283,
|
751 |
+
"logits/rejected": 2.216254472732544,
|
752 |
+
"logps/chosen": -0.887627124786377,
|
753 |
+
"logps/rejected": -1.0360788106918335,
|
754 |
+
"loss": 1.6259,
|
755 |
+
"rewards/accuracies": 0.5249999761581421,
|
756 |
+
"rewards/chosen": -2.2190678119659424,
|
757 |
+
"rewards/margins": 0.3711293637752533,
|
758 |
+
"rewards/rejected": -2.5901970863342285,
|
759 |
+
"step": 245
|
760 |
+
},
|
761 |
+
{
|
762 |
+
"epoch": 0.5343307507347048,
|
763 |
+
"grad_norm": 10.042684286997996,
|
764 |
+
"learning_rate": 2.63083989060736e-07,
|
765 |
+
"logits/chosen": 2.162696361541748,
|
766 |
+
"logits/rejected": 2.178349494934082,
|
767 |
+
"logps/chosen": -0.9883350133895874,
|
768 |
+
"logps/rejected": -1.1994340419769287,
|
769 |
+
"loss": 1.6589,
|
770 |
+
"rewards/accuracies": 0.625,
|
771 |
+
"rewards/chosen": -2.4708378314971924,
|
772 |
+
"rewards/margins": 0.5277475118637085,
|
773 |
+
"rewards/rejected": -2.9985852241516113,
|
774 |
+
"step": 250
|
775 |
+
},
|
776 |
+
{
|
777 |
+
"epoch": 0.5450173657493989,
|
778 |
+
"grad_norm": 10.520900876521566,
|
779 |
+
"learning_rate": 2.537398517538159e-07,
|
780 |
+
"logits/chosen": 2.2908027172088623,
|
781 |
+
"logits/rejected": 2.235353708267212,
|
782 |
+
"logps/chosen": -0.9431841969490051,
|
783 |
+
"logps/rejected": -1.015700101852417,
|
784 |
+
"loss": 1.6741,
|
785 |
+
"rewards/accuracies": 0.5249999761581421,
|
786 |
+
"rewards/chosen": -2.3579602241516113,
|
787 |
+
"rewards/margins": 0.18129023909568787,
|
788 |
+
"rewards/rejected": -2.539250612258911,
|
789 |
+
"step": 255
|
790 |
+
},
|
791 |
+
{
|
792 |
+
"epoch": 0.555703980764093,
|
793 |
+
"grad_norm": 9.313778052567603,
|
794 |
+
"learning_rate": 2.4439048392604877e-07,
|
795 |
+
"logits/chosen": 2.1532225608825684,
|
796 |
+
"logits/rejected": 2.2637267112731934,
|
797 |
+
"logps/chosen": -0.9131342768669128,
|
798 |
+
"logps/rejected": -1.2010549306869507,
|
799 |
+
"loss": 1.5779,
|
800 |
+
"rewards/accuracies": 0.699999988079071,
|
801 |
+
"rewards/chosen": -2.2828357219696045,
|
802 |
+
"rewards/margins": 0.7198014259338379,
|
803 |
+
"rewards/rejected": -3.0026373863220215,
|
804 |
+
"step": 260
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"epoch": 0.566390595778787,
|
808 |
+
"grad_norm": 17.127300405069718,
|
809 |
+
"learning_rate": 2.3504896151374144e-07,
|
810 |
+
"logits/chosen": 2.2980525493621826,
|
811 |
+
"logits/rejected": 2.2274222373962402,
|
812 |
+
"logps/chosen": -0.9766885042190552,
|
813 |
+
"logps/rejected": -1.0872061252593994,
|
814 |
+
"loss": 1.6741,
|
815 |
+
"rewards/accuracies": 0.550000011920929,
|
816 |
+
"rewards/chosen": -2.4417214393615723,
|
817 |
+
"rewards/margins": 0.2762938141822815,
|
818 |
+
"rewards/rejected": -2.718015193939209,
|
819 |
+
"step": 265
|
820 |
+
},
|
821 |
+
{
|
822 |
+
"epoch": 0.5770772107934812,
|
823 |
+
"grad_norm": 14.096322545468484,
|
824 |
+
"learning_rate": 2.2572834948067795e-07,
|
825 |
+
"logits/chosen": 2.13866925239563,
|
826 |
+
"logits/rejected": 2.09075665473938,
|
827 |
+
"logps/chosen": -0.866176962852478,
|
828 |
+
"logps/rejected": -1.0079139471054077,
|
829 |
+
"loss": 1.6445,
|
830 |
+
"rewards/accuracies": 0.5,
|
831 |
+
"rewards/chosen": -2.165442705154419,
|
832 |
+
"rewards/margins": 0.3543422520160675,
|
833 |
+
"rewards/rejected": -2.519784688949585,
|
834 |
+
"step": 270
|
835 |
+
},
|
836 |
+
{
|
837 |
+
"epoch": 0.5877638258081752,
|
838 |
+
"grad_norm": 11.519350142557252,
|
839 |
+
"learning_rate": 2.164416835455862e-07,
|
840 |
+
"logits/chosen": 1.9599807262420654,
|
841 |
+
"logits/rejected": 1.9166367053985596,
|
842 |
+
"logps/chosen": -0.9016801714897156,
|
843 |
+
"logps/rejected": -1.060001015663147,
|
844 |
+
"loss": 1.6308,
|
845 |
+
"rewards/accuracies": 0.5249999761581421,
|
846 |
+
"rewards/chosen": -2.2542004585266113,
|
847 |
+
"rewards/margins": 0.39580219984054565,
|
848 |
+
"rewards/rejected": -2.6500024795532227,
|
849 |
+
"step": 275
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 0.5984504408228694,
|
853 |
+
"grad_norm": 9.94425188158257,
|
854 |
+
"learning_rate": 2.072019519505062e-07,
|
855 |
+
"logits/chosen": 2.006683349609375,
|
856 |
+
"logits/rejected": 1.9877275228500366,
|
857 |
+
"logps/chosen": -0.9040509462356567,
|
858 |
+
"logps/rejected": -1.0043412446975708,
|
859 |
+
"loss": 1.6201,
|
860 |
+
"rewards/accuracies": 0.512499988079071,
|
861 |
+
"rewards/chosen": -2.260127305984497,
|
862 |
+
"rewards/margins": 0.2507258355617523,
|
863 |
+
"rewards/rejected": -2.5108530521392822,
|
864 |
+
"step": 280
|
865 |
+
},
|
866 |
+
{
|
867 |
+
"epoch": 0.6091370558375635,
|
868 |
+
"grad_norm": 17.126362799060438,
|
869 |
+
"learning_rate": 1.980220772955602e-07,
|
870 |
+
"logits/chosen": 2.040419816970825,
|
871 |
+
"logits/rejected": 1.9795433282852173,
|
872 |
+
"logps/chosen": -1.0711162090301514,
|
873 |
+
"logps/rejected": -1.1054927110671997,
|
874 |
+
"loss": 1.6696,
|
875 |
+
"rewards/accuracies": 0.48750001192092896,
|
876 |
+
"rewards/chosen": -2.677790641784668,
|
877 |
+
"rewards/margins": 0.08594076335430145,
|
878 |
+
"rewards/rejected": -2.7637314796447754,
|
879 |
+
"step": 285
|
880 |
+
},
|
881 |
+
{
|
882 |
+
"epoch": 0.6198236708522575,
|
883 |
+
"grad_norm": 13.492215975630849,
|
884 |
+
"learning_rate": 1.8891489846552644e-07,
|
885 |
+
"logits/chosen": 2.3427670001983643,
|
886 |
+
"logits/rejected": 2.3189902305603027,
|
887 |
+
"logps/chosen": -0.9356976747512817,
|
888 |
+
"logps/rejected": -1.0561234951019287,
|
889 |
+
"loss": 1.6313,
|
890 |
+
"rewards/accuracies": 0.5249999761581421,
|
891 |
+
"rewards/chosen": -2.3392443656921387,
|
892 |
+
"rewards/margins": 0.30106478929519653,
|
893 |
+
"rewards/rejected": -2.6403088569641113,
|
894 |
+
"step": 290
|
895 |
+
},
|
896 |
+
{
|
897 |
+
"epoch": 0.6305102858669517,
|
898 |
+
"grad_norm": 6.893975837082491,
|
899 |
+
"learning_rate": 1.7989315267349933e-07,
|
900 |
+
"logits/chosen": 2.3545644283294678,
|
901 |
+
"logits/rejected": 2.3726115226745605,
|
902 |
+
"logps/chosen": -0.9937087893486023,
|
903 |
+
"logps/rejected": -1.0849647521972656,
|
904 |
+
"loss": 1.6816,
|
905 |
+
"rewards/accuracies": 0.574999988079071,
|
906 |
+
"rewards/chosen": -2.484271764755249,
|
907 |
+
"rewards/margins": 0.22813992202281952,
|
908 |
+
"rewards/rejected": -2.712411642074585,
|
909 |
+
"step": 295
|
910 |
+
},
|
911 |
+
{
|
912 |
+
"epoch": 0.6411969008816457,
|
913 |
+
"grad_norm": 6.6767839699246405,
|
914 |
+
"learning_rate": 1.7096945764674398e-07,
|
915 |
+
"logits/chosen": 2.0134410858154297,
|
916 |
+
"logits/rejected": 1.9672603607177734,
|
917 |
+
"logps/chosen": -1.038081169128418,
|
918 |
+
"logps/rejected": -1.0271114110946655,
|
919 |
+
"loss": 1.6598,
|
920 |
+
"rewards/accuracies": 0.4375,
|
921 |
+
"rewards/chosen": -2.595202922821045,
|
922 |
+
"rewards/margins": -0.027424562722444534,
|
923 |
+
"rewards/rejected": -2.5677783489227295,
|
924 |
+
"step": 300
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"epoch": 0.6518835158963399,
|
928 |
+
"grad_norm": 8.402566753074327,
|
929 |
+
"learning_rate": 1.621562939796643e-07,
|
930 |
+
"logits/chosen": 2.099931240081787,
|
931 |
+
"logits/rejected": 2.131023406982422,
|
932 |
+
"logps/chosen": -0.9801462888717651,
|
933 |
+
"logps/rejected": -1.1142746210098267,
|
934 |
+
"loss": 1.6239,
|
935 |
+
"rewards/accuracies": 0.4749999940395355,
|
936 |
+
"rewards/chosen": -2.4503657817840576,
|
937 |
+
"rewards/margins": 0.33532071113586426,
|
938 |
+
"rewards/rejected": -2.785686492919922,
|
939 |
+
"step": 305
|
940 |
+
},
|
941 |
+
{
|
942 |
+
"epoch": 0.6625701309110339,
|
943 |
+
"grad_norm": 13.484008579562406,
|
944 |
+
"learning_rate": 1.5346598767856345e-07,
|
945 |
+
"logits/chosen": 2.3738818168640137,
|
946 |
+
"logits/rejected": 2.2915918827056885,
|
947 |
+
"logps/chosen": -0.9194884300231934,
|
948 |
+
"logps/rejected": -1.1123372316360474,
|
949 |
+
"loss": 1.6509,
|
950 |
+
"rewards/accuracies": 0.612500011920929,
|
951 |
+
"rewards/chosen": -2.2987208366394043,
|
952 |
+
"rewards/margins": 0.4821222722530365,
|
953 |
+
"rewards/rejected": -2.7808432579040527,
|
954 |
+
"step": 310
|
955 |
+
},
|
956 |
+
{
|
957 |
+
"epoch": 0.673256745925728,
|
958 |
+
"grad_norm": 11.276062854222866,
|
959 |
+
"learning_rate": 1.4491069292260866e-07,
|
960 |
+
"logits/chosen": 2.0688905715942383,
|
961 |
+
"logits/rejected": 2.0706334114074707,
|
962 |
+
"logps/chosen": -0.9416360855102539,
|
963 |
+
"logps/rejected": -0.9889582395553589,
|
964 |
+
"loss": 1.6263,
|
965 |
+
"rewards/accuracies": 0.48750001192092896,
|
966 |
+
"rewards/chosen": -2.354090452194214,
|
967 |
+
"rewards/margins": 0.11830528825521469,
|
968 |
+
"rewards/rejected": -2.472395658493042,
|
969 |
+
"step": 315
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 0.6839433609404221,
|
973 |
+
"grad_norm": 12.406779732367644,
|
974 |
+
"learning_rate": 1.365023750651133e-07,
|
975 |
+
"logits/chosen": 2.029076337814331,
|
976 |
+
"logits/rejected": 2.07525634765625,
|
977 |
+
"logps/chosen": -0.8669158220291138,
|
978 |
+
"logps/rejected": -1.1443169116973877,
|
979 |
+
"loss": 1.5861,
|
980 |
+
"rewards/accuracies": 0.625,
|
981 |
+
"rewards/chosen": -2.1672892570495605,
|
982 |
+
"rewards/margins": 0.6935025453567505,
|
983 |
+
"rewards/rejected": -2.8607916831970215,
|
984 |
+
"step": 320
|
985 |
+
},
|
986 |
+
{
|
987 |
+
"epoch": 0.6946299759551162,
|
988 |
+
"grad_norm": 9.410463016483124,
|
989 |
+
"learning_rate": 1.2825279389890818e-07,
|
990 |
+
"logits/chosen": 2.2680742740631104,
|
991 |
+
"logits/rejected": 2.2626852989196777,
|
992 |
+
"logps/chosen": -0.9127257466316223,
|
993 |
+
"logps/rejected": -1.036811113357544,
|
994 |
+
"loss": 1.5807,
|
995 |
+
"rewards/accuracies": 0.550000011920929,
|
996 |
+
"rewards/chosen": -2.2818140983581543,
|
997 |
+
"rewards/margins": 0.31021344661712646,
|
998 |
+
"rewards/rejected": -2.5920276641845703,
|
999 |
+
"step": 325
|
1000 |
+
},
|
1001 |
+
{
|
1002 |
+
"epoch": 0.7053165909698104,
|
1003 |
+
"grad_norm": 13.015902838409264,
|
1004 |
+
"learning_rate": 1.201734872092077e-07,
|
1005 |
+
"logits/chosen": 2.0204429626464844,
|
1006 |
+
"logits/rejected": 1.9889049530029297,
|
1007 |
+
"logps/chosen": -0.9624814987182617,
|
1008 |
+
"logps/rejected": -1.4148118495941162,
|
1009 |
+
"loss": 1.6247,
|
1010 |
+
"rewards/accuracies": 0.550000011920929,
|
1011 |
+
"rewards/chosen": -2.4062037467956543,
|
1012 |
+
"rewards/margins": 1.130826473236084,
|
1013 |
+
"rewards/rejected": -3.537029981613159,
|
1014 |
+
"step": 330
|
1015 |
+
},
|
1016 |
+
{
|
1017 |
+
"epoch": 0.7160032059845044,
|
1018 |
+
"grad_norm": 10.457412577224535,
|
1019 |
+
"learning_rate": 1.1227575463697439e-07,
|
1020 |
+
"logits/chosen": 2.1002488136291504,
|
1021 |
+
"logits/rejected": 2.1578989028930664,
|
1022 |
+
"logps/chosen": -0.9219167828559875,
|
1023 |
+
"logps/rejected": -1.0541021823883057,
|
1024 |
+
"loss": 1.6126,
|
1025 |
+
"rewards/accuracies": 0.574999988079071,
|
1026 |
+
"rewards/chosen": -2.3047919273376465,
|
1027 |
+
"rewards/margins": 0.33046382665634155,
|
1028 |
+
"rewards/rejected": -2.635255813598633,
|
1029 |
+
"step": 335
|
1030 |
+
},
|
1031 |
+
{
|
1032 |
+
"epoch": 0.7266898209991985,
|
1033 |
+
"grad_norm": 13.23573306912224,
|
1034 |
+
"learning_rate": 1.0457064187534861e-07,
|
1035 |
+
"logits/chosen": 2.2117228507995605,
|
1036 |
+
"logits/rejected": 2.291081666946411,
|
1037 |
+
"logps/chosen": -0.8333222270011902,
|
1038 |
+
"logps/rejected": -1.0662933588027954,
|
1039 |
+
"loss": 1.611,
|
1040 |
+
"rewards/accuracies": 0.574999988079071,
|
1041 |
+
"rewards/chosen": -2.083305597305298,
|
1042 |
+
"rewards/margins": 0.5824280977249146,
|
1043 |
+
"rewards/rejected": -2.665733575820923,
|
1044 |
+
"step": 340
|
1045 |
+
},
|
1046 |
+
{
|
1047 |
+
"epoch": 0.7373764360138926,
|
1048 |
+
"grad_norm": 7.317920448430426,
|
1049 |
+
"learning_rate": 9.706892522124838e-08,
|
1050 |
+
"logits/chosen": 2.1216635704040527,
|
1051 |
+
"logits/rejected": 2.178093194961548,
|
1052 |
+
"logps/chosen": -1.0588462352752686,
|
1053 |
+
"logps/rejected": -1.0970056056976318,
|
1054 |
+
"loss": 1.5811,
|
1055 |
+
"rewards/accuracies": 0.512499988079071,
|
1056 |
+
"rewards/chosen": -2.647115468978882,
|
1057 |
+
"rewards/margins": 0.09539877623319626,
|
1058 |
+
"rewards/rejected": -2.742514133453369,
|
1059 |
+
"step": 345
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 0.7480630510285867,
|
1063 |
+
"grad_norm": 14.193162676817792,
|
1064 |
+
"learning_rate": 8.978109650374396e-08,
|
1065 |
+
"logits/chosen": 2.2546029090881348,
|
1066 |
+
"logits/rejected": 2.294689893722534,
|
1067 |
+
"logps/chosen": -0.9023507833480835,
|
1068 |
+
"logps/rejected": -1.162570595741272,
|
1069 |
+
"loss": 1.6171,
|
1070 |
+
"rewards/accuracies": 0.5874999761581421,
|
1071 |
+
"rewards/chosen": -2.2558767795562744,
|
1072 |
+
"rewards/margins": 0.6505497694015503,
|
1073 |
+
"rewards/rejected": -2.9064266681671143,
|
1074 |
+
"step": 350
|
1075 |
+
},
|
1076 |
+
{
|
1077 |
+
"epoch": 0.7587496660432808,
|
1078 |
+
"grad_norm": 12.52863776377329,
|
1079 |
+
"learning_rate": 8.271734841028552e-08,
|
1080 |
+
"logits/chosen": 2.245272636413574,
|
1081 |
+
"logits/rejected": 2.2576751708984375,
|
1082 |
+
"logps/chosen": -1.005053997039795,
|
1083 |
+
"logps/rejected": -1.0007752180099487,
|
1084 |
+
"loss": 1.6536,
|
1085 |
+
"rewards/accuracies": 0.4749999940395355,
|
1086 |
+
"rewards/chosen": -2.512634754180908,
|
1087 |
+
"rewards/margins": -0.010696577839553356,
|
1088 |
+
"rewards/rejected": -2.5019383430480957,
|
1089 |
+
"step": 355
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 0.7694362810579749,
|
1093 |
+
"grad_norm": 10.369189534811127,
|
1094 |
+
"learning_rate": 7.588756023130833e-08,
|
1095 |
+
"logits/chosen": 2.318972587585449,
|
1096 |
+
"logits/rejected": 2.281419515609741,
|
1097 |
+
"logps/chosen": -1.0578750371932983,
|
1098 |
+
"logps/rejected": -1.3604905605316162,
|
1099 |
+
"loss": 1.6205,
|
1100 |
+
"rewards/accuracies": 0.625,
|
1101 |
+
"rewards/chosen": -2.6446874141693115,
|
1102 |
+
"rewards/margins": 0.7565388679504395,
|
1103 |
+
"rewards/rejected": -3.40122652053833,
|
1104 |
+
"step": 360
|
1105 |
+
},
|
1106 |
+
{
|
1107 |
+
"epoch": 0.7801228960726689,
|
1108 |
+
"grad_norm": 6.075365195709392,
|
1109 |
+
"learning_rate": 6.930128404315214e-08,
|
1110 |
+
"logits/chosen": 2.0289673805236816,
|
1111 |
+
"logits/rejected": 2.0680978298187256,
|
1112 |
+
"logps/chosen": -0.976359486579895,
|
1113 |
+
"logps/rejected": -1.0074145793914795,
|
1114 |
+
"loss": 1.6246,
|
1115 |
+
"rewards/accuracies": 0.512499988079071,
|
1116 |
+
"rewards/chosen": -2.440898895263672,
|
1117 |
+
"rewards/margins": 0.07763750851154327,
|
1118 |
+
"rewards/rejected": -2.518536329269409,
|
1119 |
+
"step": 365
|
1120 |
+
},
|
1121 |
+
{
|
1122 |
+
"epoch": 0.7908095110873631,
|
1123 |
+
"grad_norm": 16.214823251732284,
|
1124 |
+
"learning_rate": 6.296773134861824e-08,
|
1125 |
+
"logits/chosen": 2.2002553939819336,
|
1126 |
+
"logits/rejected": 2.2343714237213135,
|
1127 |
+
"logps/chosen": -0.9719416499137878,
|
1128 |
+
"logps/rejected": -1.023166537284851,
|
1129 |
+
"loss": 1.6894,
|
1130 |
+
"rewards/accuracies": 0.5874999761581421,
|
1131 |
+
"rewards/chosen": -2.429853916168213,
|
1132 |
+
"rewards/margins": 0.12806224822998047,
|
1133 |
+
"rewards/rejected": -2.5579161643981934,
|
1134 |
+
"step": 370
|
1135 |
+
},
|
1136 |
+
{
|
1137 |
+
"epoch": 0.8014961261020572,
|
1138 |
+
"grad_norm": 10.840080143128308,
|
1139 |
+
"learning_rate": 5.6895760193850145e-08,
|
1140 |
+
"logits/chosen": 2.1092734336853027,
|
1141 |
+
"logits/rejected": 2.168839693069458,
|
1142 |
+
"logps/chosen": -1.0623986721038818,
|
1143 |
+
"logps/rejected": -1.1893246173858643,
|
1144 |
+
"loss": 1.6248,
|
1145 |
+
"rewards/accuracies": 0.675000011920929,
|
1146 |
+
"rewards/chosen": -2.6559970378875732,
|
1147 |
+
"rewards/margins": 0.31731492280960083,
|
1148 |
+
"rewards/rejected": -2.9733119010925293,
|
1149 |
+
"step": 375
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 0.8121827411167513,
|
1153 |
+
"grad_norm": 14.986128731201607,
|
1154 |
+
"learning_rate": 5.109386277955477e-08,
|
1155 |
+
"logits/chosen": 2.400944232940674,
|
1156 |
+
"logits/rejected": 2.424126625061035,
|
1157 |
+
"logps/chosen": -1.0006974935531616,
|
1158 |
+
"logps/rejected": -1.0918558835983276,
|
1159 |
+
"loss": 1.6152,
|
1160 |
+
"rewards/accuracies": 0.48750001192092896,
|
1161 |
+
"rewards/chosen": -2.501743793487549,
|
1162 |
+
"rewards/margins": 0.22789612412452698,
|
1163 |
+
"rewards/rejected": -2.729639768600464,
|
1164 |
+
"step": 380
|
1165 |
+
},
|
1166 |
+
{
|
1167 |
+
"epoch": 0.8228693561314454,
|
1168 |
+
"grad_norm": 11.04509407630343,
|
1169 |
+
"learning_rate": 4.557015358389216e-08,
|
1170 |
+
"logits/chosen": 2.2161147594451904,
|
1171 |
+
"logits/rejected": 2.255049705505371,
|
1172 |
+
"logps/chosen": -0.9763911962509155,
|
1173 |
+
"logps/rejected": -1.2085295915603638,
|
1174 |
+
"loss": 1.6246,
|
1175 |
+
"rewards/accuracies": 0.5625,
|
1176 |
+
"rewards/chosen": -2.4409780502319336,
|
1177 |
+
"rewards/margins": 0.5803458094596863,
|
1178 |
+
"rewards/rejected": -3.0213241577148438,
|
1179 |
+
"step": 385
|
1180 |
+
},
|
1181 |
+
{
|
1182 |
+
"epoch": 0.8335559711461394,
|
1183 |
+
"grad_norm": 17.08993723906078,
|
1184 |
+
"learning_rate": 4.0332358013644015e-08,
|
1185 |
+
"logits/chosen": 2.3219351768493652,
|
1186 |
+
"logits/rejected": 2.316385507583618,
|
1187 |
+
"logps/chosen": -0.8949235081672668,
|
1188 |
+
"logps/rejected": -1.2672293186187744,
|
1189 |
+
"loss": 1.6161,
|
1190 |
+
"rewards/accuracies": 0.5375000238418579,
|
1191 |
+
"rewards/chosen": -2.2373085021972656,
|
1192 |
+
"rewards/margins": 0.9307647943496704,
|
1193 |
+
"rewards/rejected": -3.1680736541748047,
|
1194 |
+
"step": 390
|
1195 |
+
},
|
1196 |
+
{
|
1197 |
+
"epoch": 0.8442425861608336,
|
1198 |
+
"grad_norm": 7.486937950213667,
|
1199 |
+
"learning_rate": 3.538780159953347e-08,
|
1200 |
+
"logits/chosen": 2.171618700027466,
|
1201 |
+
"logits/rejected": 2.133547067642212,
|
1202 |
+
"logps/chosen": -0.913979172706604,
|
1203 |
+
"logps/rejected": -1.011819839477539,
|
1204 |
+
"loss": 1.6306,
|
1205 |
+
"rewards/accuracies": 0.512499988079071,
|
1206 |
+
"rewards/chosen": -2.284947633743286,
|
1207 |
+
"rewards/margins": 0.24460192024707794,
|
1208 |
+
"rewards/rejected": -2.5295493602752686,
|
1209 |
+
"step": 395
|
1210 |
+
},
|
1211 |
+
{
|
1212 |
+
"epoch": 0.8549292011755276,
|
1213 |
+
"grad_norm": 12.731226626352624,
|
1214 |
+
"learning_rate": 3.074339975080836e-08,
|
1215 |
+
"logits/chosen": 2.207425594329834,
|
1216 |
+
"logits/rejected": 2.2549290657043457,
|
1217 |
+
"logps/chosen": -1.030874490737915,
|
1218 |
+
"logps/rejected": -1.2776422500610352,
|
1219 |
+
"loss": 1.6417,
|
1220 |
+
"rewards/accuracies": 0.5,
|
1221 |
+
"rewards/chosen": -2.577186107635498,
|
1222 |
+
"rewards/margins": 0.6169193983078003,
|
1223 |
+
"rewards/rejected": -3.194105863571167,
|
1224 |
+
"step": 400
|
1225 |
+
},
|
1226 |
+
{
|
1227 |
+
"epoch": 0.8549292011755276,
|
1228 |
+
"eval_logits/chosen": 2.0145604610443115,
|
1229 |
+
"eval_logits/rejected": 2.0190441608428955,
|
1230 |
+
"eval_logps/chosen": -0.8955881595611572,
|
1231 |
+
"eval_logps/rejected": -1.058284044265747,
|
1232 |
+
"eval_loss": 1.6235809326171875,
|
1233 |
+
"eval_rewards/accuracies": 0.6209677457809448,
|
1234 |
+
"eval_rewards/chosen": -2.2389702796936035,
|
1235 |
+
"eval_rewards/margins": 0.40673962235450745,
|
1236 |
+
"eval_rewards/rejected": -2.645709753036499,
|
1237 |
+
"eval_runtime": 44.336,
|
1238 |
+
"eval_samples_per_second": 44.23,
|
1239 |
+
"eval_steps_per_second": 1.398,
|
1240 |
+
"step": 400
|
1241 |
+
},
|
1242 |
+
{
|
1243 |
+
"epoch": 0.8656158161902218,
|
1244 |
+
"grad_norm": 10.867554141726604,
|
1245 |
+
"learning_rate": 2.6405648083415833e-08,
|
1246 |
+
"logits/chosen": 2.151744842529297,
|
1247 |
+
"logits/rejected": 2.200650691986084,
|
1248 |
+
"logps/chosen": -0.980237603187561,
|
1249 |
+
"logps/rejected": -1.0543644428253174,
|
1250 |
+
"loss": 1.7482,
|
1251 |
+
"rewards/accuracies": 0.5249999761581421,
|
1252 |
+
"rewards/chosen": -2.450594186782837,
|
1253 |
+
"rewards/margins": 0.18531720340251923,
|
1254 |
+
"rewards/rejected": -2.635911226272583,
|
1255 |
+
"step": 405
|
1256 |
+
},
|
1257 |
+
{
|
1258 |
+
"epoch": 0.8763024312049158,
|
1259 |
+
"grad_norm": 8.929503879536968,
|
1260 |
+
"learning_rate": 2.2380613335296033e-08,
|
1261 |
+
"logits/chosen": 2.186063289642334,
|
1262 |
+
"logits/rejected": 2.2187790870666504,
|
1263 |
+
"logps/chosen": -0.9089058041572571,
|
1264 |
+
"logps/rejected": -1.0911632776260376,
|
1265 |
+
"loss": 1.5847,
|
1266 |
+
"rewards/accuracies": 0.612500011920929,
|
1267 |
+
"rewards/chosen": -2.272264242172241,
|
1268 |
+
"rewards/margins": 0.45564380288124084,
|
1269 |
+
"rewards/rejected": -2.727908134460449,
|
1270 |
+
"step": 410
|
1271 |
+
},
|
1272 |
+
{
|
1273 |
+
"epoch": 0.88698904621961,
|
1274 |
+
"grad_norm": 7.217288770833436,
|
1275 |
+
"learning_rate": 1.8673924881500823e-08,
|
1276 |
+
"logits/chosen": 2.505709409713745,
|
1277 |
+
"logits/rejected": 2.470183849334717,
|
1278 |
+
"logps/chosen": -0.8387972116470337,
|
1279 |
+
"logps/rejected": -1.0221149921417236,
|
1280 |
+
"loss": 1.6029,
|
1281 |
+
"rewards/accuracies": 0.5,
|
1282 |
+
"rewards/chosen": -2.0969929695129395,
|
1283 |
+
"rewards/margins": 0.45829471945762634,
|
1284 |
+
"rewards/rejected": -2.5552875995635986,
|
1285 |
+
"step": 415
|
1286 |
+
},
|
1287 |
+
{
|
1288 |
+
"epoch": 0.897675661234304,
|
1289 |
+
"grad_norm": 8.941360081202905,
|
1290 |
+
"learning_rate": 1.5290766861003475e-08,
|
1291 |
+
"logits/chosen": 2.191061019897461,
|
1292 |
+
"logits/rejected": 2.173102378845215,
|
1293 |
+
"logps/chosen": -0.9252120852470398,
|
1294 |
+
"logps/rejected": -1.03108811378479,
|
1295 |
+
"loss": 1.6027,
|
1296 |
+
"rewards/accuracies": 0.550000011920929,
|
1297 |
+
"rewards/chosen": -2.31303071975708,
|
1298 |
+
"rewards/margins": 0.2646896243095398,
|
1299 |
+
"rewards/rejected": -2.5777201652526855,
|
1300 |
+
"step": 420
|
1301 |
+
},
|
1302 |
+
{
|
1303 |
+
"epoch": 0.9083622762489981,
|
1304 |
+
"grad_norm": 14.568857335011407,
|
1305 |
+
"learning_rate": 1.2235870926211616e-08,
|
1306 |
+
"logits/chosen": 2.0643928050994873,
|
1307 |
+
"logits/rejected": 2.0896859169006348,
|
1308 |
+
"logps/chosen": -0.9114057421684265,
|
1309 |
+
"logps/rejected": -1.0619531869888306,
|
1310 |
+
"loss": 1.6274,
|
1311 |
+
"rewards/accuracies": 0.5375000238418579,
|
1312 |
+
"rewards/chosen": -2.2785143852233887,
|
1313 |
+
"rewards/margins": 0.37636831402778625,
|
1314 |
+
"rewards/rejected": -2.6548826694488525,
|
1315 |
+
"step": 425
|
1316 |
+
},
|
1317 |
+
{
|
1318 |
+
"epoch": 0.9190488912636923,
|
1319 |
+
"grad_norm": 12.761707033542287,
|
1320 |
+
"learning_rate": 9.513509625323518e-09,
|
1321 |
+
"logits/chosen": 2.1255576610565186,
|
1322 |
+
"logits/rejected": 2.181107759475708,
|
1323 |
+
"logps/chosen": -0.85987389087677,
|
1324 |
+
"logps/rejected": -1.0060930252075195,
|
1325 |
+
"loss": 1.5736,
|
1326 |
+
"rewards/accuracies": 0.7124999761581421,
|
1327 |
+
"rewards/chosen": -2.149684429168701,
|
1328 |
+
"rewards/margins": 0.3655479848384857,
|
1329 |
+
"rewards/rejected": -2.515232563018799,
|
1330 |
+
"step": 430
|
1331 |
+
},
|
1332 |
+
{
|
1333 |
+
"epoch": 0.9297355062783863,
|
1334 |
+
"grad_norm": 16.68383920608418,
|
1335 |
+
"learning_rate": 7.127490426783123e-09,
|
1336 |
+
"logits/chosen": 2.2676453590393066,
|
1337 |
+
"logits/rejected": 2.224987506866455,
|
1338 |
+
"logps/chosen": -0.9531320333480835,
|
1339 |
+
"logps/rejected": -1.1150267124176025,
|
1340 |
+
"loss": 1.6026,
|
1341 |
+
"rewards/accuracies": 0.574999988079071,
|
1342 |
+
"rewards/chosen": -2.3828301429748535,
|
1343 |
+
"rewards/margins": 0.4047369956970215,
|
1344 |
+
"rewards/rejected": -2.787567138671875,
|
1345 |
+
"step": 435
|
1346 |
+
},
|
1347 |
+
{
|
1348 |
+
"epoch": 0.9404221212930804,
|
1349 |
+
"grad_norm": 11.429700658399854,
|
1350 |
+
"learning_rate": 5.08115039419113e-09,
|
1351 |
+
"logits/chosen": 2.2096710205078125,
|
1352 |
+
"logits/rejected": 2.2330925464630127,
|
1353 |
+
"logps/chosen": -0.8609305620193481,
|
1354 |
+
"logps/rejected": -0.9673101305961609,
|
1355 |
+
"loss": 1.5978,
|
1356 |
+
"rewards/accuracies": 0.574999988079071,
|
1357 |
+
"rewards/chosen": -2.1523263454437256,
|
1358 |
+
"rewards/margins": 0.265949010848999,
|
1359 |
+
"rewards/rejected": -2.4182753562927246,
|
1360 |
+
"step": 440
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 0.9511087363077745,
|
1364 |
+
"grad_norm": 7.426544303240404,
|
1365 |
+
"learning_rate": 3.3773515191196646e-09,
|
1366 |
+
"logits/chosen": 2.20963716506958,
|
1367 |
+
"logits/rejected": 2.201622724533081,
|
1368 |
+
"logps/chosen": -0.8218202590942383,
|
1369 |
+
"logps/rejected": -0.8649258613586426,
|
1370 |
+
"loss": 1.6414,
|
1371 |
+
"rewards/accuracies": 0.512499988079071,
|
1372 |
+
"rewards/chosen": -2.0545506477355957,
|
1373 |
+
"rewards/margins": 0.10776408016681671,
|
1374 |
+
"rewards/rejected": -2.1623148918151855,
|
1375 |
+
"step": 445
|
1376 |
+
},
|
1377 |
+
{
|
1378 |
+
"epoch": 0.9617953513224686,
|
1379 |
+
"grad_norm": 12.349821491663924,
|
1380 |
+
"learning_rate": 2.0184767183584474e-09,
|
1381 |
+
"logits/chosen": 2.325861930847168,
|
1382 |
+
"logits/rejected": 2.2664694786071777,
|
1383 |
+
"logps/chosen": -0.9952027201652527,
|
1384 |
+
"logps/rejected": -1.3025867938995361,
|
1385 |
+
"loss": 1.6015,
|
1386 |
+
"rewards/accuracies": 0.5874999761581421,
|
1387 |
+
"rewards/chosen": -2.488007068634033,
|
1388 |
+
"rewards/margins": 0.7684597969055176,
|
1389 |
+
"rewards/rejected": -3.2564666271209717,
|
1390 |
+
"step": 450
|
1391 |
+
},
|
1392 |
+
{
|
1393 |
+
"epoch": 0.9724819663371627,
|
1394 |
+
"grad_norm": 10.41544260252462,
|
1395 |
+
"learning_rate": 1.0064265011902328e-09,
|
1396 |
+
"logits/chosen": 2.0219030380249023,
|
1397 |
+
"logits/rejected": 1.9868360757827759,
|
1398 |
+
"logps/chosen": -1.0134050846099854,
|
1399 |
+
"logps/rejected": -1.0333858728408813,
|
1400 |
+
"loss": 1.6377,
|
1401 |
+
"rewards/accuracies": 0.4000000059604645,
|
1402 |
+
"rewards/chosen": -2.533512592315674,
|
1403 |
+
"rewards/margins": 0.049951858818531036,
|
1404 |
+
"rewards/rejected": -2.5834643840789795,
|
1405 |
+
"step": 455
|
1406 |
+
},
|
1407 |
+
{
|
1408 |
+
"epoch": 0.9831685813518568,
|
1409 |
+
"grad_norm": 9.044209064035641,
|
1410 |
+
"learning_rate": 3.4261631135654167e-10,
|
1411 |
+
"logits/chosen": 2.2376327514648438,
|
1412 |
+
"logits/rejected": 2.1572484970092773,
|
1413 |
+
"logps/chosen": -1.0239425897598267,
|
1414 |
+
"logps/rejected": -1.3165332078933716,
|
1415 |
+
"loss": 1.5875,
|
1416 |
+
"rewards/accuracies": 0.5375000238418579,
|
1417 |
+
"rewards/chosen": -2.559856414794922,
|
1418 |
+
"rewards/margins": 0.7314767241477966,
|
1419 |
+
"rewards/rejected": -3.291332960128784,
|
1420 |
+
"step": 460
|
1421 |
+
},
|
1422 |
+
{
|
1423 |
+
"epoch": 0.9938551963665508,
|
1424 |
+
"grad_norm": 11.3294414057108,
|
1425 |
+
"learning_rate": 2.797454743164174e-11,
|
1426 |
+
"logits/chosen": 2.186420202255249,
|
1427 |
+
"logits/rejected": 2.1705265045166016,
|
1428 |
+
"logps/chosen": -0.875818133354187,
|
1429 |
+
"logps/rejected": -1.0201952457427979,
|
1430 |
+
"loss": 1.5967,
|
1431 |
+
"rewards/accuracies": 0.5249999761581421,
|
1432 |
+
"rewards/chosen": -2.1895456314086914,
|
1433 |
+
"rewards/margins": 0.36094242334365845,
|
1434 |
+
"rewards/rejected": -2.550488233566284,
|
1435 |
+
"step": 465
|
1436 |
+
},
|
1437 |
+
{
|
1438 |
+
"epoch": 0.9981298423724285,
|
1439 |
+
"step": 467,
|
1440 |
+
"total_flos": 0.0,
|
1441 |
+
"train_loss": 1.6566840494470372,
|
1442 |
+
"train_runtime": 5417.7619,
|
1443 |
+
"train_samples_per_second": 11.052,
|
1444 |
+
"train_steps_per_second": 0.086
|
1445 |
+
}
|
1446 |
+
],
|
1447 |
+
"logging_steps": 5,
|
1448 |
+
"max_steps": 467,
|
1449 |
+
"num_input_tokens_seen": 0,
|
1450 |
+
"num_train_epochs": 1,
|
1451 |
+
"save_steps": 1000000,
|
1452 |
+
"stateful_callbacks": {
|
1453 |
+
"TrainerControl": {
|
1454 |
+
"args": {
|
1455 |
+
"should_epoch_stop": false,
|
1456 |
+
"should_evaluate": false,
|
1457 |
+
"should_log": false,
|
1458 |
+
"should_save": false,
|
1459 |
+
"should_training_stop": false
|
1460 |
+
},
|
1461 |
+
"attributes": {}
|
1462 |
+
}
|
1463 |
+
},
|
1464 |
+
"total_flos": 0.0,
|
1465 |
+
"train_batch_size": 2,
|
1466 |
+
"trial_name": null,
|
1467 |
+
"trial_params": null
|
1468 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f2b77a893b9f093dd80df1a7919ac514a78023dfb6581fd6cfc149f1fd2b6d92
|
3 |
+
size 6584
|