juyongjiang
commited on
Commit
•
1c988ac
1
Parent(s):
1871753
upload model checkpoint
Browse files- .gitattributes +1 -0
- README.md +76 -3
- adapter_config.json +29 -0
- adapter_model.safetensors +3 -0
- all_results.json +14 -0
- config.json +44 -0
- eval_results.json +8 -0
- runs/Jun13_05-58-32_gpu1-1/events.out.tfevents.1718229608.gpu1-1.440205.0 +3 -0
- runs/Jun13_05-58-32_gpu1-1/events.out.tfevents.1718231465.gpu1-1.440205.1 +3 -0
- special_tokens_map.json +34 -0
- tokenizer.json +3 -0
- tokenizer_config.json +70 -0
- train_results.json +9 -0
- trainer_state.json +1181 -0
- training_args.bin +3 -0
.gitattributes
CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
README.md
CHANGED
@@ -1,3 +1,76 @@
|
|
1 |
-
---
|
2 |
-
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: peft
|
3 |
+
tags:
|
4 |
+
- alignment-handbook
|
5 |
+
- generated_from_trainer
|
6 |
+
datasets:
|
7 |
+
- llama-duo/synth_summarize_dataset_dedup
|
8 |
+
base_model: google/gemma-7b
|
9 |
+
model-index:
|
10 |
+
- name: gemma7b-summarize-claude3sonnet-32k
|
11 |
+
results: []
|
12 |
+
---
|
13 |
+
|
14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
15 |
+
should probably proofread and complete it, then remove this comment. -->
|
16 |
+
|
17 |
+
# gemma7b-summarize-claude3sonnet-32k
|
18 |
+
|
19 |
+
This model is a fine-tuned version of [google/gemma-7b](https://huggingface.co/google/gemma-7b) on the llama-duo/synth_summarize_dataset_dedup dataset.
|
20 |
+
It achieves the following results on the evaluation set:
|
21 |
+
- Loss: 2.5524
|
22 |
+
|
23 |
+
## Model description
|
24 |
+
|
25 |
+
More information needed
|
26 |
+
|
27 |
+
## Intended uses & limitations
|
28 |
+
|
29 |
+
More information needed
|
30 |
+
|
31 |
+
## Training and evaluation data
|
32 |
+
|
33 |
+
More information needed
|
34 |
+
|
35 |
+
## Training procedure
|
36 |
+
|
37 |
+
### Training hyperparameters
|
38 |
+
|
39 |
+
The following hyperparameters were used during training:
|
40 |
+
- learning_rate: 0.0002
|
41 |
+
- train_batch_size: 4
|
42 |
+
- eval_batch_size: 2
|
43 |
+
- seed: 42
|
44 |
+
- distributed_type: multi-GPU
|
45 |
+
- num_devices: 8
|
46 |
+
- gradient_accumulation_steps: 2
|
47 |
+
- total_train_batch_size: 64
|
48 |
+
- total_eval_batch_size: 16
|
49 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
50 |
+
- lr_scheduler_type: cosine
|
51 |
+
- lr_scheduler_warmup_ratio: 0.1
|
52 |
+
- num_epochs: 10
|
53 |
+
|
54 |
+
### Training results
|
55 |
+
|
56 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
57 |
+
|:-------------:|:-----:|:----:|:---------------:|
|
58 |
+
| 1.6901 | 1.0 | 76 | 2.9966 |
|
59 |
+
| 1.1272 | 2.0 | 152 | 2.6070 |
|
60 |
+
| 1.0337 | 3.0 | 228 | 2.5657 |
|
61 |
+
| 0.9638 | 4.0 | 304 | 2.5379 |
|
62 |
+
| 0.9419 | 5.0 | 380 | 2.5376 |
|
63 |
+
| 0.9117 | 6.0 | 456 | 2.5333 |
|
64 |
+
| 0.8944 | 7.0 | 532 | 2.5417 |
|
65 |
+
| 0.8824 | 8.0 | 608 | 2.5474 |
|
66 |
+
| 0.8759 | 9.0 | 684 | 2.5541 |
|
67 |
+
| 0.8735 | 10.0 | 760 | 2.5524 |
|
68 |
+
|
69 |
+
|
70 |
+
### Framework versions
|
71 |
+
|
72 |
+
- PEFT 0.10.0
|
73 |
+
- Transformers 4.40.0
|
74 |
+
- Pytorch 2.1.2+cu121
|
75 |
+
- Datasets 2.18.0
|
76 |
+
- Tokenizers 0.19.1
|
adapter_config.json
ADDED
@@ -0,0 +1,29 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"alpha_pattern": {},
|
3 |
+
"auto_mapping": null,
|
4 |
+
"base_model_name_or_path": "google/gemma-7b",
|
5 |
+
"bias": "none",
|
6 |
+
"fan_in_fan_out": false,
|
7 |
+
"inference_mode": true,
|
8 |
+
"init_lora_weights": true,
|
9 |
+
"layer_replication": null,
|
10 |
+
"layers_pattern": null,
|
11 |
+
"layers_to_transform": null,
|
12 |
+
"loftq_config": {},
|
13 |
+
"lora_alpha": 32,
|
14 |
+
"lora_dropout": 0.05,
|
15 |
+
"megatron_config": null,
|
16 |
+
"megatron_core": "megatron.core",
|
17 |
+
"modules_to_save": null,
|
18 |
+
"peft_type": "LORA",
|
19 |
+
"r": 16,
|
20 |
+
"rank_pattern": {},
|
21 |
+
"revision": null,
|
22 |
+
"target_modules": [
|
23 |
+
"v_proj",
|
24 |
+
"q_proj"
|
25 |
+
],
|
26 |
+
"task_type": "CAUSAL_LM",
|
27 |
+
"use_dora": false,
|
28 |
+
"use_rslora": false
|
29 |
+
}
|
adapter_model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6246c0c074e47dbab1c260de365a1dd3cd5f6dee4b8edc86131bd452b7492d8f
|
3 |
+
size 12860096
|
all_results.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"eval_loss": 2.5523786544799805,
|
4 |
+
"eval_runtime": 0.2402,
|
5 |
+
"eval_samples": 25,
|
6 |
+
"eval_samples_per_second": 41.632,
|
7 |
+
"eval_steps_per_second": 4.163,
|
8 |
+
"total_flos": 2.318334860390826e+18,
|
9 |
+
"train_loss": 2.2621336485210217,
|
10 |
+
"train_runtime": 1856.2046,
|
11 |
+
"train_samples": 31747,
|
12 |
+
"train_samples_per_second": 26.156,
|
13 |
+
"train_steps_per_second": 0.409
|
14 |
+
}
|
config.json
ADDED
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "google/gemma-7b",
|
3 |
+
"architectures": [
|
4 |
+
"GemmaForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_bias": false,
|
7 |
+
"attention_dropout": 0.0,
|
8 |
+
"bos_token_id": 2,
|
9 |
+
"eos_token_id": 1,
|
10 |
+
"head_dim": 256,
|
11 |
+
"hidden_act": "gelu",
|
12 |
+
"hidden_activation": null,
|
13 |
+
"hidden_size": 3072,
|
14 |
+
"initializer_range": 0.02,
|
15 |
+
"intermediate_size": 24576,
|
16 |
+
"max_position_embeddings": 8192,
|
17 |
+
"model_type": "gemma",
|
18 |
+
"num_attention_heads": 16,
|
19 |
+
"num_hidden_layers": 28,
|
20 |
+
"num_key_value_heads": 16,
|
21 |
+
"pad_token_id": 0,
|
22 |
+
"quantization_config": {
|
23 |
+
"_load_in_4bit": true,
|
24 |
+
"_load_in_8bit": false,
|
25 |
+
"bnb_4bit_compute_dtype": "bfloat16",
|
26 |
+
"bnb_4bit_quant_storage": "uint8",
|
27 |
+
"bnb_4bit_quant_type": "nf4",
|
28 |
+
"bnb_4bit_use_double_quant": false,
|
29 |
+
"llm_int8_enable_fp32_cpu_offload": false,
|
30 |
+
"llm_int8_has_fp16_weight": false,
|
31 |
+
"llm_int8_skip_modules": null,
|
32 |
+
"llm_int8_threshold": 6.0,
|
33 |
+
"load_in_4bit": true,
|
34 |
+
"load_in_8bit": false,
|
35 |
+
"quant_method": "bitsandbytes"
|
36 |
+
},
|
37 |
+
"rms_norm_eps": 1e-06,
|
38 |
+
"rope_scaling": null,
|
39 |
+
"rope_theta": 10000.0,
|
40 |
+
"torch_dtype": "bfloat16",
|
41 |
+
"transformers_version": "4.40.0",
|
42 |
+
"use_cache": true,
|
43 |
+
"vocab_size": 256000
|
44 |
+
}
|
eval_results.json
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"eval_loss": 2.5523786544799805,
|
4 |
+
"eval_runtime": 0.2402,
|
5 |
+
"eval_samples": 25,
|
6 |
+
"eval_samples_per_second": 41.632,
|
7 |
+
"eval_steps_per_second": 4.163
|
8 |
+
}
|
runs/Jun13_05-58-32_gpu1-1/events.out.tfevents.1718229608.gpu1-1.440205.0
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:496d8947cbc337a0c498ea8911e5c7c0774ac3699b879cee1d7140ea5726b3cf
|
3 |
+
size 40721
|
runs/Jun13_05-58-32_gpu1-1/events.out.tfevents.1718231465.gpu1-1.440205.1
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b23bca1b61bda1a02121cfeb428edc004551a46a8dde1ba53809071c871f1e94
|
3 |
+
size 359
|
special_tokens_map.json
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"<|im_start|>",
|
4 |
+
"<|im_end|>"
|
5 |
+
],
|
6 |
+
"bos_token": {
|
7 |
+
"content": "<bos>",
|
8 |
+
"lstrip": false,
|
9 |
+
"normalized": false,
|
10 |
+
"rstrip": false,
|
11 |
+
"single_word": false
|
12 |
+
},
|
13 |
+
"eos_token": {
|
14 |
+
"content": "<eos>",
|
15 |
+
"lstrip": false,
|
16 |
+
"normalized": false,
|
17 |
+
"rstrip": false,
|
18 |
+
"single_word": false
|
19 |
+
},
|
20 |
+
"pad_token": {
|
21 |
+
"content": "<pad>",
|
22 |
+
"lstrip": false,
|
23 |
+
"normalized": false,
|
24 |
+
"rstrip": false,
|
25 |
+
"single_word": false
|
26 |
+
},
|
27 |
+
"unk_token": {
|
28 |
+
"content": "<unk>",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": false,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false
|
33 |
+
}
|
34 |
+
}
|
tokenizer.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:322a5f52ab5cab196761ab397a022d6fa3a2e1418585e532bb6efb2fedd2ae94
|
3 |
+
size 17477501
|
tokenizer_config.json
ADDED
@@ -0,0 +1,70 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"add_bos_token": false,
|
3 |
+
"add_eos_token": false,
|
4 |
+
"added_tokens_decoder": {
|
5 |
+
"0": {
|
6 |
+
"content": "<pad>",
|
7 |
+
"lstrip": false,
|
8 |
+
"normalized": false,
|
9 |
+
"rstrip": false,
|
10 |
+
"single_word": false,
|
11 |
+
"special": true
|
12 |
+
},
|
13 |
+
"1": {
|
14 |
+
"content": "<eos>",
|
15 |
+
"lstrip": false,
|
16 |
+
"normalized": false,
|
17 |
+
"rstrip": false,
|
18 |
+
"single_word": false,
|
19 |
+
"special": true
|
20 |
+
},
|
21 |
+
"2": {
|
22 |
+
"content": "<bos>",
|
23 |
+
"lstrip": false,
|
24 |
+
"normalized": false,
|
25 |
+
"rstrip": false,
|
26 |
+
"single_word": false,
|
27 |
+
"special": true
|
28 |
+
},
|
29 |
+
"3": {
|
30 |
+
"content": "<unk>",
|
31 |
+
"lstrip": false,
|
32 |
+
"normalized": false,
|
33 |
+
"rstrip": false,
|
34 |
+
"single_word": false,
|
35 |
+
"special": true
|
36 |
+
},
|
37 |
+
"106": {
|
38 |
+
"content": "<|im_start|>",
|
39 |
+
"lstrip": false,
|
40 |
+
"normalized": false,
|
41 |
+
"rstrip": false,
|
42 |
+
"single_word": false,
|
43 |
+
"special": true
|
44 |
+
},
|
45 |
+
"107": {
|
46 |
+
"content": "<|im_end|>",
|
47 |
+
"lstrip": false,
|
48 |
+
"normalized": false,
|
49 |
+
"rstrip": false,
|
50 |
+
"single_word": false,
|
51 |
+
"special": true
|
52 |
+
}
|
53 |
+
},
|
54 |
+
"additional_special_tokens": [
|
55 |
+
"<|im_start|>",
|
56 |
+
"<|im_end|>"
|
57 |
+
],
|
58 |
+
"bos_token": "<bos>",
|
59 |
+
"chat_template": "{% if messages[0]['role'] == 'user' or messages[0]['role'] == 'system' %}{{ bos_token }}{% endif %}{% for message in messages %}{{ '<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n' }}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% elif messages[-1]['role'] == 'assistant' %}{{ eos_token }}{% endif %}",
|
60 |
+
"clean_up_tokenization_spaces": false,
|
61 |
+
"eos_token": "<eos>",
|
62 |
+
"legacy": null,
|
63 |
+
"model_max_length": 2048,
|
64 |
+
"pad_token": "<pad>",
|
65 |
+
"sp_model_kwargs": {},
|
66 |
+
"spaces_between_special_tokens": false,
|
67 |
+
"tokenizer_class": "GemmaTokenizer",
|
68 |
+
"unk_token": "<unk>",
|
69 |
+
"use_default_system_prompt": false
|
70 |
+
}
|
train_results.json
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"epoch": 10.0,
|
3 |
+
"total_flos": 2.318334860390826e+18,
|
4 |
+
"train_loss": 2.2621336485210217,
|
5 |
+
"train_runtime": 1856.2046,
|
6 |
+
"train_samples": 31747,
|
7 |
+
"train_samples_per_second": 26.156,
|
8 |
+
"train_steps_per_second": 0.409
|
9 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,1181 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 760,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.013157894736842105,
|
13 |
+
"grad_norm": 251.0,
|
14 |
+
"learning_rate": 2.631578947368421e-06,
|
15 |
+
"loss": 35.372,
|
16 |
+
"step": 1
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.06578947368421052,
|
20 |
+
"grad_norm": 239.0,
|
21 |
+
"learning_rate": 1.3157894736842106e-05,
|
22 |
+
"loss": 35.3482,
|
23 |
+
"step": 5
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.13157894736842105,
|
27 |
+
"grad_norm": 83.5,
|
28 |
+
"learning_rate": 2.6315789473684212e-05,
|
29 |
+
"loss": 30.5543,
|
30 |
+
"step": 10
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.19736842105263158,
|
34 |
+
"grad_norm": 37.5,
|
35 |
+
"learning_rate": 3.9473684210526316e-05,
|
36 |
+
"loss": 23.0935,
|
37 |
+
"step": 15
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.2631578947368421,
|
41 |
+
"grad_norm": 17.875,
|
42 |
+
"learning_rate": 5.2631578947368424e-05,
|
43 |
+
"loss": 18.9215,
|
44 |
+
"step": 20
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.32894736842105265,
|
48 |
+
"grad_norm": 11.125,
|
49 |
+
"learning_rate": 6.578947368421054e-05,
|
50 |
+
"loss": 16.7379,
|
51 |
+
"step": 25
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.39473684210526316,
|
55 |
+
"grad_norm": 3.671875,
|
56 |
+
"learning_rate": 7.894736842105263e-05,
|
57 |
+
"loss": 14.9012,
|
58 |
+
"step": 30
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.4605263157894737,
|
62 |
+
"grad_norm": 3.09375,
|
63 |
+
"learning_rate": 9.210526315789474e-05,
|
64 |
+
"loss": 14.2412,
|
65 |
+
"step": 35
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.5263157894736842,
|
69 |
+
"grad_norm": 4.0,
|
70 |
+
"learning_rate": 0.00010526315789473685,
|
71 |
+
"loss": 13.5992,
|
72 |
+
"step": 40
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.5921052631578947,
|
76 |
+
"grad_norm": 6.15625,
|
77 |
+
"learning_rate": 0.00011842105263157894,
|
78 |
+
"loss": 12.7389,
|
79 |
+
"step": 45
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.6578947368421053,
|
83 |
+
"grad_norm": 11.875,
|
84 |
+
"learning_rate": 0.00013157894736842108,
|
85 |
+
"loss": 11.4753,
|
86 |
+
"step": 50
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.7236842105263158,
|
90 |
+
"grad_norm": 19.25,
|
91 |
+
"learning_rate": 0.00014473684210526317,
|
92 |
+
"loss": 8.7885,
|
93 |
+
"step": 55
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.7894736842105263,
|
97 |
+
"grad_norm": 20.75,
|
98 |
+
"learning_rate": 0.00015789473684210527,
|
99 |
+
"loss": 4.8341,
|
100 |
+
"step": 60
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.8552631578947368,
|
104 |
+
"grad_norm": 5.90625,
|
105 |
+
"learning_rate": 0.00017105263157894739,
|
106 |
+
"loss": 2.389,
|
107 |
+
"step": 65
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.9210526315789473,
|
111 |
+
"grad_norm": 3.953125,
|
112 |
+
"learning_rate": 0.00018421052631578948,
|
113 |
+
"loss": 1.9682,
|
114 |
+
"step": 70
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.9868421052631579,
|
118 |
+
"grad_norm": 1.078125,
|
119 |
+
"learning_rate": 0.00019736842105263157,
|
120 |
+
"loss": 1.6901,
|
121 |
+
"step": 75
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 1.0,
|
125 |
+
"eval_loss": 2.9966416358947754,
|
126 |
+
"eval_runtime": 0.2381,
|
127 |
+
"eval_samples_per_second": 41.997,
|
128 |
+
"eval_steps_per_second": 4.2,
|
129 |
+
"step": 76
|
130 |
+
},
|
131 |
+
{
|
132 |
+
"epoch": 1.0526315789473684,
|
133 |
+
"grad_norm": 1.0234375,
|
134 |
+
"learning_rate": 0.00019998312416333227,
|
135 |
+
"loss": 1.565,
|
136 |
+
"step": 80
|
137 |
+
},
|
138 |
+
{
|
139 |
+
"epoch": 1.118421052631579,
|
140 |
+
"grad_norm": 0.765625,
|
141 |
+
"learning_rate": 0.0001999145758387301,
|
142 |
+
"loss": 1.4481,
|
143 |
+
"step": 85
|
144 |
+
},
|
145 |
+
{
|
146 |
+
"epoch": 1.1842105263157894,
|
147 |
+
"grad_norm": 0.97265625,
|
148 |
+
"learning_rate": 0.00019979333640833947,
|
149 |
+
"loss": 1.3767,
|
150 |
+
"step": 90
|
151 |
+
},
|
152 |
+
{
|
153 |
+
"epoch": 1.25,
|
154 |
+
"grad_norm": 1.078125,
|
155 |
+
"learning_rate": 0.00019961946980917456,
|
156 |
+
"loss": 1.3246,
|
157 |
+
"step": 95
|
158 |
+
},
|
159 |
+
{
|
160 |
+
"epoch": 1.3157894736842106,
|
161 |
+
"grad_norm": 0.58203125,
|
162 |
+
"learning_rate": 0.00019939306773179497,
|
163 |
+
"loss": 1.2875,
|
164 |
+
"step": 100
|
165 |
+
},
|
166 |
+
{
|
167 |
+
"epoch": 1.381578947368421,
|
168 |
+
"grad_norm": 0.46484375,
|
169 |
+
"learning_rate": 0.00019911424957195158,
|
170 |
+
"loss": 1.244,
|
171 |
+
"step": 105
|
172 |
+
},
|
173 |
+
{
|
174 |
+
"epoch": 1.4473684210526316,
|
175 |
+
"grad_norm": 0.859375,
|
176 |
+
"learning_rate": 0.00019878316236762196,
|
177 |
+
"loss": 1.2335,
|
178 |
+
"step": 110
|
179 |
+
},
|
180 |
+
{
|
181 |
+
"epoch": 1.513157894736842,
|
182 |
+
"grad_norm": 0.78515625,
|
183 |
+
"learning_rate": 0.000198399980721468,
|
184 |
+
"loss": 1.2052,
|
185 |
+
"step": 115
|
186 |
+
},
|
187 |
+
{
|
188 |
+
"epoch": 1.5789473684210527,
|
189 |
+
"grad_norm": 0.59765625,
|
190 |
+
"learning_rate": 0.0001979649067087574,
|
191 |
+
"loss": 1.1933,
|
192 |
+
"step": 120
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 1.6447368421052633,
|
196 |
+
"grad_norm": 1.0703125,
|
197 |
+
"learning_rate": 0.00019747816977079671,
|
198 |
+
"loss": 1.1718,
|
199 |
+
"step": 125
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 1.7105263157894737,
|
203 |
+
"grad_norm": 1.7734375,
|
204 |
+
"learning_rate": 0.00019694002659393305,
|
205 |
+
"loss": 1.1603,
|
206 |
+
"step": 130
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 1.776315789473684,
|
210 |
+
"grad_norm": 0.5703125,
|
211 |
+
"learning_rate": 0.00019635076097418734,
|
212 |
+
"loss": 1.1664,
|
213 |
+
"step": 135
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.8421052631578947,
|
217 |
+
"grad_norm": 0.765625,
|
218 |
+
"learning_rate": 0.00019571068366759143,
|
219 |
+
"loss": 1.1523,
|
220 |
+
"step": 140
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 1.9078947368421053,
|
224 |
+
"grad_norm": 0.51953125,
|
225 |
+
"learning_rate": 0.00019502013222630712,
|
226 |
+
"loss": 1.1372,
|
227 |
+
"step": 145
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 1.973684210526316,
|
231 |
+
"grad_norm": 0.703125,
|
232 |
+
"learning_rate": 0.00019427947082061432,
|
233 |
+
"loss": 1.1272,
|
234 |
+
"step": 150
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 2.0,
|
238 |
+
"eval_loss": 2.6070284843444824,
|
239 |
+
"eval_runtime": 0.2379,
|
240 |
+
"eval_samples_per_second": 42.033,
|
241 |
+
"eval_steps_per_second": 4.203,
|
242 |
+
"step": 152
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 2.039473684210526,
|
246 |
+
"grad_norm": 0.6875,
|
247 |
+
"learning_rate": 0.00019348909004686152,
|
248 |
+
"loss": 1.1013,
|
249 |
+
"step": 155
|
250 |
+
},
|
251 |
+
{
|
252 |
+
"epoch": 2.1052631578947367,
|
253 |
+
"grad_norm": 1.15625,
|
254 |
+
"learning_rate": 0.00019264940672148018,
|
255 |
+
"loss": 1.1011,
|
256 |
+
"step": 160
|
257 |
+
},
|
258 |
+
{
|
259 |
+
"epoch": 2.1710526315789473,
|
260 |
+
"grad_norm": 0.59765625,
|
261 |
+
"learning_rate": 0.00019176086366117211,
|
262 |
+
"loss": 1.0822,
|
263 |
+
"step": 165
|
264 |
+
},
|
265 |
+
{
|
266 |
+
"epoch": 2.236842105263158,
|
267 |
+
"grad_norm": 1.21875,
|
268 |
+
"learning_rate": 0.00019082392944938466,
|
269 |
+
"loss": 1.0853,
|
270 |
+
"step": 170
|
271 |
+
},
|
272 |
+
{
|
273 |
+
"epoch": 2.3026315789473686,
|
274 |
+
"grad_norm": 0.859375,
|
275 |
+
"learning_rate": 0.0001898390981891979,
|
276 |
+
"loss": 1.0778,
|
277 |
+
"step": 175
|
278 |
+
},
|
279 |
+
{
|
280 |
+
"epoch": 2.3684210526315788,
|
281 |
+
"grad_norm": 0.58203125,
|
282 |
+
"learning_rate": 0.00018880688924275378,
|
283 |
+
"loss": 1.0601,
|
284 |
+
"step": 180
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 2.4342105263157894,
|
288 |
+
"grad_norm": 0.8359375,
|
289 |
+
"learning_rate": 0.0001877278469573643,
|
290 |
+
"loss": 1.0645,
|
291 |
+
"step": 185
|
292 |
+
},
|
293 |
+
{
|
294 |
+
"epoch": 2.5,
|
295 |
+
"grad_norm": 1.0078125,
|
296 |
+
"learning_rate": 0.00018660254037844388,
|
297 |
+
"loss": 1.0663,
|
298 |
+
"step": 190
|
299 |
+
},
|
300 |
+
{
|
301 |
+
"epoch": 2.5657894736842106,
|
302 |
+
"grad_norm": 2.015625,
|
303 |
+
"learning_rate": 0.0001854315629494165,
|
304 |
+
"loss": 1.0582,
|
305 |
+
"step": 195
|
306 |
+
},
|
307 |
+
{
|
308 |
+
"epoch": 2.6315789473684212,
|
309 |
+
"grad_norm": 1.015625,
|
310 |
+
"learning_rate": 0.00018421553219875658,
|
311 |
+
"loss": 1.05,
|
312 |
+
"step": 200
|
313 |
+
},
|
314 |
+
{
|
315 |
+
"epoch": 2.6973684210526314,
|
316 |
+
"grad_norm": 2.375,
|
317 |
+
"learning_rate": 0.00018295508941432815,
|
318 |
+
"loss": 1.0463,
|
319 |
+
"step": 205
|
320 |
+
},
|
321 |
+
{
|
322 |
+
"epoch": 2.763157894736842,
|
323 |
+
"grad_norm": 0.7890625,
|
324 |
+
"learning_rate": 0.0001816508993051943,
|
325 |
+
"loss": 1.0336,
|
326 |
+
"step": 210
|
327 |
+
},
|
328 |
+
{
|
329 |
+
"epoch": 2.8289473684210527,
|
330 |
+
"grad_norm": 0.84375,
|
331 |
+
"learning_rate": 0.0001803036496510752,
|
332 |
+
"loss": 1.0233,
|
333 |
+
"step": 215
|
334 |
+
},
|
335 |
+
{
|
336 |
+
"epoch": 2.8947368421052633,
|
337 |
+
"grad_norm": 0.76171875,
|
338 |
+
"learning_rate": 0.00017891405093963938,
|
339 |
+
"loss": 1.0287,
|
340 |
+
"step": 220
|
341 |
+
},
|
342 |
+
{
|
343 |
+
"epoch": 2.9605263157894735,
|
344 |
+
"grad_norm": 0.75390625,
|
345 |
+
"learning_rate": 0.00017748283599182014,
|
346 |
+
"loss": 1.0337,
|
347 |
+
"step": 225
|
348 |
+
},
|
349 |
+
{
|
350 |
+
"epoch": 3.0,
|
351 |
+
"eval_loss": 2.565699338912964,
|
352 |
+
"eval_runtime": 0.2367,
|
353 |
+
"eval_samples_per_second": 42.254,
|
354 |
+
"eval_steps_per_second": 4.225,
|
355 |
+
"step": 228
|
356 |
+
},
|
357 |
+
{
|
358 |
+
"epoch": 3.026315789473684,
|
359 |
+
"grad_norm": 0.9375,
|
360 |
+
"learning_rate": 0.00017601075957535364,
|
361 |
+
"loss": 1.0152,
|
362 |
+
"step": 230
|
363 |
+
},
|
364 |
+
{
|
365 |
+
"epoch": 3.0921052631578947,
|
366 |
+
"grad_norm": 1.109375,
|
367 |
+
"learning_rate": 0.00017449859800674371,
|
368 |
+
"loss": 0.9987,
|
369 |
+
"step": 235
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 3.1578947368421053,
|
373 |
+
"grad_norm": 2.046875,
|
374 |
+
"learning_rate": 0.0001729471487418621,
|
375 |
+
"loss": 0.9947,
|
376 |
+
"step": 240
|
377 |
+
},
|
378 |
+
{
|
379 |
+
"epoch": 3.223684210526316,
|
380 |
+
"grad_norm": 0.94921875,
|
381 |
+
"learning_rate": 0.00017135722995540107,
|
382 |
+
"loss": 0.9919,
|
383 |
+
"step": 245
|
384 |
+
},
|
385 |
+
{
|
386 |
+
"epoch": 3.2894736842105265,
|
387 |
+
"grad_norm": 0.99609375,
|
388 |
+
"learning_rate": 0.00016972968010939954,
|
389 |
+
"loss": 0.9945,
|
390 |
+
"step": 250
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"epoch": 3.3552631578947367,
|
394 |
+
"grad_norm": 0.890625,
|
395 |
+
"learning_rate": 0.00016806535751107037,
|
396 |
+
"loss": 0.9904,
|
397 |
+
"step": 255
|
398 |
+
},
|
399 |
+
{
|
400 |
+
"epoch": 3.4210526315789473,
|
401 |
+
"grad_norm": 1.1171875,
|
402 |
+
"learning_rate": 0.00016636513986016213,
|
403 |
+
"loss": 0.9886,
|
404 |
+
"step": 260
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 3.486842105263158,
|
408 |
+
"grad_norm": 1.1171875,
|
409 |
+
"learning_rate": 0.00016462992378609407,
|
410 |
+
"loss": 0.99,
|
411 |
+
"step": 265
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 3.5526315789473686,
|
415 |
+
"grad_norm": 0.78515625,
|
416 |
+
"learning_rate": 0.0001628606243751082,
|
417 |
+
"loss": 0.9947,
|
418 |
+
"step": 270
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"epoch": 3.6184210526315788,
|
422 |
+
"grad_norm": 0.76953125,
|
423 |
+
"learning_rate": 0.00016105817468768798,
|
424 |
+
"loss": 0.991,
|
425 |
+
"step": 275
|
426 |
+
},
|
427 |
+
{
|
428 |
+
"epoch": 3.6842105263157894,
|
429 |
+
"grad_norm": 1.03125,
|
430 |
+
"learning_rate": 0.00015922352526649803,
|
431 |
+
"loss": 0.982,
|
432 |
+
"step": 280
|
433 |
+
},
|
434 |
+
{
|
435 |
+
"epoch": 3.75,
|
436 |
+
"grad_norm": 1.1171875,
|
437 |
+
"learning_rate": 0.0001573576436351046,
|
438 |
+
"loss": 0.9797,
|
439 |
+
"step": 285
|
440 |
+
},
|
441 |
+
{
|
442 |
+
"epoch": 3.8157894736842106,
|
443 |
+
"grad_norm": 1.703125,
|
444 |
+
"learning_rate": 0.00015546151378774086,
|
445 |
+
"loss": 0.9724,
|
446 |
+
"step": 290
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"epoch": 3.8815789473684212,
|
450 |
+
"grad_norm": 0.6796875,
|
451 |
+
"learning_rate": 0.00015353613567038607,
|
452 |
+
"loss": 0.9695,
|
453 |
+
"step": 295
|
454 |
+
},
|
455 |
+
{
|
456 |
+
"epoch": 3.9473684210526314,
|
457 |
+
"grad_norm": 0.87109375,
|
458 |
+
"learning_rate": 0.00015158252465343242,
|
459 |
+
"loss": 0.9638,
|
460 |
+
"step": 300
|
461 |
+
},
|
462 |
+
{
|
463 |
+
"epoch": 4.0,
|
464 |
+
"eval_loss": 2.537924289703369,
|
465 |
+
"eval_runtime": 0.2365,
|
466 |
+
"eval_samples_per_second": 42.288,
|
467 |
+
"eval_steps_per_second": 4.229,
|
468 |
+
"step": 304
|
469 |
+
},
|
470 |
+
{
|
471 |
+
"epoch": 4.0131578947368425,
|
472 |
+
"grad_norm": 0.7734375,
|
473 |
+
"learning_rate": 0.00014960171099621795,
|
474 |
+
"loss": 0.9669,
|
475 |
+
"step": 305
|
476 |
+
},
|
477 |
+
{
|
478 |
+
"epoch": 4.078947368421052,
|
479 |
+
"grad_norm": 0.7109375,
|
480 |
+
"learning_rate": 0.00014759473930370736,
|
481 |
+
"loss": 0.9414,
|
482 |
+
"step": 310
|
483 |
+
},
|
484 |
+
{
|
485 |
+
"epoch": 4.144736842105263,
|
486 |
+
"grad_norm": 0.921875,
|
487 |
+
"learning_rate": 0.00014556266797560732,
|
488 |
+
"loss": 0.9589,
|
489 |
+
"step": 315
|
490 |
+
},
|
491 |
+
{
|
492 |
+
"epoch": 4.2105263157894735,
|
493 |
+
"grad_norm": 0.66796875,
|
494 |
+
"learning_rate": 0.00014350656864820733,
|
495 |
+
"loss": 0.9489,
|
496 |
+
"step": 320
|
497 |
+
},
|
498 |
+
{
|
499 |
+
"epoch": 4.276315789473684,
|
500 |
+
"grad_norm": 0.79296875,
|
501 |
+
"learning_rate": 0.00014142752562923988,
|
502 |
+
"loss": 0.9502,
|
503 |
+
"step": 325
|
504 |
+
},
|
505 |
+
{
|
506 |
+
"epoch": 4.342105263157895,
|
507 |
+
"grad_norm": 1.2578125,
|
508 |
+
"learning_rate": 0.0001393266353260583,
|
509 |
+
"loss": 0.9626,
|
510 |
+
"step": 330
|
511 |
+
},
|
512 |
+
{
|
513 |
+
"epoch": 4.407894736842105,
|
514 |
+
"grad_norm": 0.59765625,
|
515 |
+
"learning_rate": 0.00013720500566743362,
|
516 |
+
"loss": 0.9511,
|
517 |
+
"step": 335
|
518 |
+
},
|
519 |
+
{
|
520 |
+
"epoch": 4.473684210526316,
|
521 |
+
"grad_norm": 1.65625,
|
522 |
+
"learning_rate": 0.00013506375551927547,
|
523 |
+
"loss": 0.9456,
|
524 |
+
"step": 340
|
525 |
+
},
|
526 |
+
{
|
527 |
+
"epoch": 4.5394736842105265,
|
528 |
+
"grad_norm": 1.4765625,
|
529 |
+
"learning_rate": 0.00013290401409458532,
|
530 |
+
"loss": 0.9461,
|
531 |
+
"step": 345
|
532 |
+
},
|
533 |
+
{
|
534 |
+
"epoch": 4.605263157894737,
|
535 |
+
"grad_norm": 1.875,
|
536 |
+
"learning_rate": 0.00013072692035795305,
|
537 |
+
"loss": 0.9418,
|
538 |
+
"step": 350
|
539 |
+
},
|
540 |
+
{
|
541 |
+
"epoch": 4.671052631578947,
|
542 |
+
"grad_norm": 0.98828125,
|
543 |
+
"learning_rate": 0.00012853362242491053,
|
544 |
+
"loss": 0.9481,
|
545 |
+
"step": 355
|
546 |
+
},
|
547 |
+
{
|
548 |
+
"epoch": 4.7368421052631575,
|
549 |
+
"grad_norm": 0.87890625,
|
550 |
+
"learning_rate": 0.00012632527695645993,
|
551 |
+
"loss": 0.9461,
|
552 |
+
"step": 360
|
553 |
+
},
|
554 |
+
{
|
555 |
+
"epoch": 4.802631578947368,
|
556 |
+
"grad_norm": 1.140625,
|
557 |
+
"learning_rate": 0.00012410304854909495,
|
558 |
+
"loss": 0.9493,
|
559 |
+
"step": 365
|
560 |
+
},
|
561 |
+
{
|
562 |
+
"epoch": 4.868421052631579,
|
563 |
+
"grad_norm": 0.87890625,
|
564 |
+
"learning_rate": 0.0001218681091206376,
|
565 |
+
"loss": 0.9283,
|
566 |
+
"step": 370
|
567 |
+
},
|
568 |
+
{
|
569 |
+
"epoch": 4.934210526315789,
|
570 |
+
"grad_norm": 0.62109375,
|
571 |
+
"learning_rate": 0.0001196216372922136,
|
572 |
+
"loss": 0.9395,
|
573 |
+
"step": 375
|
574 |
+
},
|
575 |
+
{
|
576 |
+
"epoch": 5.0,
|
577 |
+
"grad_norm": 1.6484375,
|
578 |
+
"learning_rate": 0.00011736481776669306,
|
579 |
+
"loss": 0.9419,
|
580 |
+
"step": 380
|
581 |
+
},
|
582 |
+
{
|
583 |
+
"epoch": 5.0,
|
584 |
+
"eval_loss": 2.5376482009887695,
|
585 |
+
"eval_runtime": 0.2349,
|
586 |
+
"eval_samples_per_second": 42.567,
|
587 |
+
"eval_steps_per_second": 4.257,
|
588 |
+
"step": 380
|
589 |
+
},
|
590 |
+
{
|
591 |
+
"epoch": 5.065789473684211,
|
592 |
+
"grad_norm": 0.95703125,
|
593 |
+
"learning_rate": 0.00011509884070392369,
|
594 |
+
"loss": 0.9265,
|
595 |
+
"step": 385
|
596 |
+
},
|
597 |
+
{
|
598 |
+
"epoch": 5.131578947368421,
|
599 |
+
"grad_norm": 0.60546875,
|
600 |
+
"learning_rate": 0.00011282490109308633,
|
601 |
+
"loss": 0.9248,
|
602 |
+
"step": 390
|
603 |
+
},
|
604 |
+
{
|
605 |
+
"epoch": 5.197368421052632,
|
606 |
+
"grad_norm": 0.6484375,
|
607 |
+
"learning_rate": 0.00011054419812250338,
|
608 |
+
"loss": 0.915,
|
609 |
+
"step": 395
|
610 |
+
},
|
611 |
+
{
|
612 |
+
"epoch": 5.2631578947368425,
|
613 |
+
"grad_norm": 0.61328125,
|
614 |
+
"learning_rate": 0.00010825793454723325,
|
615 |
+
"loss": 0.9219,
|
616 |
+
"step": 400
|
617 |
+
},
|
618 |
+
{
|
619 |
+
"epoch": 5.328947368421053,
|
620 |
+
"grad_norm": 0.73046875,
|
621 |
+
"learning_rate": 0.0001059673160547834,
|
622 |
+
"loss": 0.9223,
|
623 |
+
"step": 405
|
624 |
+
},
|
625 |
+
{
|
626 |
+
"epoch": 5.394736842105263,
|
627 |
+
"grad_norm": 0.64453125,
|
628 |
+
"learning_rate": 0.00010367355062927726,
|
629 |
+
"loss": 0.9236,
|
630 |
+
"step": 410
|
631 |
+
},
|
632 |
+
{
|
633 |
+
"epoch": 5.4605263157894735,
|
634 |
+
"grad_norm": 0.6640625,
|
635 |
+
"learning_rate": 0.00010137784791440965,
|
636 |
+
"loss": 0.9252,
|
637 |
+
"step": 415
|
638 |
+
},
|
639 |
+
{
|
640 |
+
"epoch": 5.526315789473684,
|
641 |
+
"grad_norm": 0.7890625,
|
642 |
+
"learning_rate": 9.908141857552737e-05,
|
643 |
+
"loss": 0.917,
|
644 |
+
"step": 420
|
645 |
+
},
|
646 |
+
{
|
647 |
+
"epoch": 5.592105263157895,
|
648 |
+
"grad_norm": 0.671875,
|
649 |
+
"learning_rate": 9.678547366117083e-05,
|
650 |
+
"loss": 0.9213,
|
651 |
+
"step": 425
|
652 |
+
},
|
653 |
+
{
|
654 |
+
"epoch": 5.657894736842105,
|
655 |
+
"grad_norm": 0.84375,
|
656 |
+
"learning_rate": 9.449122396441345e-05,
|
657 |
+
"loss": 0.9098,
|
658 |
+
"step": 430
|
659 |
+
},
|
660 |
+
{
|
661 |
+
"epoch": 5.723684210526316,
|
662 |
+
"grad_norm": 0.8203125,
|
663 |
+
"learning_rate": 9.219987938433621e-05,
|
664 |
+
"loss": 0.9233,
|
665 |
+
"step": 435
|
666 |
+
},
|
667 |
+
{
|
668 |
+
"epoch": 5.7894736842105265,
|
669 |
+
"grad_norm": 0.77734375,
|
670 |
+
"learning_rate": 8.991264828797319e-05,
|
671 |
+
"loss": 0.9113,
|
672 |
+
"step": 440
|
673 |
+
},
|
674 |
+
{
|
675 |
+
"epoch": 5.855263157894737,
|
676 |
+
"grad_norm": 0.55859375,
|
677 |
+
"learning_rate": 8.763073687306524e-05,
|
678 |
+
"loss": 0.9198,
|
679 |
+
"step": 445
|
680 |
+
},
|
681 |
+
{
|
682 |
+
"epoch": 5.921052631578947,
|
683 |
+
"grad_norm": 1.1328125,
|
684 |
+
"learning_rate": 8.535534853195786e-05,
|
685 |
+
"loss": 0.9186,
|
686 |
+
"step": 450
|
687 |
+
},
|
688 |
+
{
|
689 |
+
"epoch": 5.9868421052631575,
|
690 |
+
"grad_norm": 0.6796875,
|
691 |
+
"learning_rate": 8.308768321697815e-05,
|
692 |
+
"loss": 0.9117,
|
693 |
+
"step": 455
|
694 |
+
},
|
695 |
+
{
|
696 |
+
"epoch": 6.0,
|
697 |
+
"eval_loss": 2.5333404541015625,
|
698 |
+
"eval_runtime": 0.2353,
|
699 |
+
"eval_samples_per_second": 42.503,
|
700 |
+
"eval_steps_per_second": 4.25,
|
701 |
+
"step": 456
|
702 |
+
},
|
703 |
+
{
|
704 |
+
"epoch": 6.052631578947368,
|
705 |
+
"grad_norm": 0.53515625,
|
706 |
+
"learning_rate": 8.082893680762619e-05,
|
707 |
+
"loss": 0.9067,
|
708 |
+
"step": 460
|
709 |
+
},
|
710 |
+
{
|
711 |
+
"epoch": 6.118421052631579,
|
712 |
+
"grad_norm": 0.73828125,
|
713 |
+
"learning_rate": 7.858030047991411e-05,
|
714 |
+
"loss": 0.9008,
|
715 |
+
"step": 465
|
716 |
+
},
|
717 |
+
{
|
718 |
+
"epoch": 6.184210526315789,
|
719 |
+
"grad_norm": 0.5546875,
|
720 |
+
"learning_rate": 7.634296007818576e-05,
|
721 |
+
"loss": 0.8891,
|
722 |
+
"step": 470
|
723 |
+
},
|
724 |
+
{
|
725 |
+
"epoch": 6.25,
|
726 |
+
"grad_norm": 0.59375,
|
727 |
+
"learning_rate": 7.411809548974792e-05,
|
728 |
+
"loss": 0.9044,
|
729 |
+
"step": 475
|
730 |
+
},
|
731 |
+
{
|
732 |
+
"epoch": 6.315789473684211,
|
733 |
+
"grad_norm": 0.53515625,
|
734 |
+
"learning_rate": 7.190688002264308e-05,
|
735 |
+
"loss": 0.8954,
|
736 |
+
"step": 480
|
737 |
+
},
|
738 |
+
{
|
739 |
+
"epoch": 6.381578947368421,
|
740 |
+
"grad_norm": 0.55859375,
|
741 |
+
"learning_rate": 6.971047978689189e-05,
|
742 |
+
"loss": 0.895,
|
743 |
+
"step": 485
|
744 |
+
},
|
745 |
+
{
|
746 |
+
"epoch": 6.447368421052632,
|
747 |
+
"grad_norm": 0.54296875,
|
748 |
+
"learning_rate": 6.753005307953167e-05,
|
749 |
+
"loss": 0.9084,
|
750 |
+
"step": 490
|
751 |
+
},
|
752 |
+
{
|
753 |
+
"epoch": 6.5131578947368425,
|
754 |
+
"grad_norm": 0.55859375,
|
755 |
+
"learning_rate": 6.536674977377496e-05,
|
756 |
+
"loss": 0.8972,
|
757 |
+
"step": 495
|
758 |
+
},
|
759 |
+
{
|
760 |
+
"epoch": 6.578947368421053,
|
761 |
+
"grad_norm": 0.66796875,
|
762 |
+
"learning_rate": 6.322171071261071e-05,
|
763 |
+
"loss": 0.8986,
|
764 |
+
"step": 500
|
765 |
+
},
|
766 |
+
{
|
767 |
+
"epoch": 6.644736842105263,
|
768 |
+
"grad_norm": 0.5234375,
|
769 |
+
"learning_rate": 6.109606710716741e-05,
|
770 |
+
"loss": 0.8889,
|
771 |
+
"step": 505
|
772 |
+
},
|
773 |
+
{
|
774 |
+
"epoch": 6.7105263157894735,
|
775 |
+
"grad_norm": 0.859375,
|
776 |
+
"learning_rate": 5.8990939940156e-05,
|
777 |
+
"loss": 0.9018,
|
778 |
+
"step": 510
|
779 |
+
},
|
780 |
+
{
|
781 |
+
"epoch": 6.776315789473684,
|
782 |
+
"grad_norm": 0.53515625,
|
783 |
+
"learning_rate": 5.690743937470657e-05,
|
784 |
+
"loss": 0.9085,
|
785 |
+
"step": 515
|
786 |
+
},
|
787 |
+
{
|
788 |
+
"epoch": 6.842105263157895,
|
789 |
+
"grad_norm": 0.640625,
|
790 |
+
"learning_rate": 5.484666416891109e-05,
|
791 |
+
"loss": 0.9082,
|
792 |
+
"step": 520
|
793 |
+
},
|
794 |
+
{
|
795 |
+
"epoch": 6.907894736842105,
|
796 |
+
"grad_norm": 0.54296875,
|
797 |
+
"learning_rate": 5.280970109638047e-05,
|
798 |
+
"loss": 0.8921,
|
799 |
+
"step": 525
|
800 |
+
},
|
801 |
+
{
|
802 |
+
"epoch": 6.973684210526316,
|
803 |
+
"grad_norm": 0.58203125,
|
804 |
+
"learning_rate": 5.079762437312219e-05,
|
805 |
+
"loss": 0.8944,
|
806 |
+
"step": 530
|
807 |
+
},
|
808 |
+
{
|
809 |
+
"epoch": 7.0,
|
810 |
+
"eval_loss": 2.5417115688323975,
|
811 |
+
"eval_runtime": 0.2355,
|
812 |
+
"eval_samples_per_second": 42.467,
|
813 |
+
"eval_steps_per_second": 4.247,
|
814 |
+
"step": 532
|
815 |
+
},
|
816 |
+
{
|
817 |
+
"epoch": 7.0394736842105265,
|
818 |
+
"grad_norm": 0.5078125,
|
819 |
+
"learning_rate": 4.8811495091039926e-05,
|
820 |
+
"loss": 0.8912,
|
821 |
+
"step": 535
|
822 |
+
},
|
823 |
+
{
|
824 |
+
"epoch": 7.105263157894737,
|
825 |
+
"grad_norm": 0.58984375,
|
826 |
+
"learning_rate": 4.685236065835443e-05,
|
827 |
+
"loss": 0.8893,
|
828 |
+
"step": 540
|
829 |
+
},
|
830 |
+
{
|
831 |
+
"epoch": 7.171052631578948,
|
832 |
+
"grad_norm": 0.5078125,
|
833 |
+
"learning_rate": 4.492125424724086e-05,
|
834 |
+
"loss": 0.8881,
|
835 |
+
"step": 545
|
836 |
+
},
|
837 |
+
{
|
838 |
+
"epoch": 7.2368421052631575,
|
839 |
+
"grad_norm": 0.5234375,
|
840 |
+
"learning_rate": 4.301919424897338e-05,
|
841 |
+
"loss": 0.8913,
|
842 |
+
"step": 550
|
843 |
+
},
|
844 |
+
{
|
845 |
+
"epoch": 7.302631578947368,
|
846 |
+
"grad_norm": 0.498046875,
|
847 |
+
"learning_rate": 4.114718373686481e-05,
|
848 |
+
"loss": 0.8871,
|
849 |
+
"step": 555
|
850 |
+
},
|
851 |
+
{
|
852 |
+
"epoch": 7.368421052631579,
|
853 |
+
"grad_norm": 0.55859375,
|
854 |
+
"learning_rate": 3.9306209937284346e-05,
|
855 |
+
"loss": 0.8854,
|
856 |
+
"step": 560
|
857 |
+
},
|
858 |
+
{
|
859 |
+
"epoch": 7.434210526315789,
|
860 |
+
"grad_norm": 0.6328125,
|
861 |
+
"learning_rate": 3.749724370903216e-05,
|
862 |
+
"loss": 0.8908,
|
863 |
+
"step": 565
|
864 |
+
},
|
865 |
+
{
|
866 |
+
"epoch": 7.5,
|
867 |
+
"grad_norm": 0.51171875,
|
868 |
+
"learning_rate": 3.5721239031346066e-05,
|
869 |
+
"loss": 0.8887,
|
870 |
+
"step": 570
|
871 |
+
},
|
872 |
+
{
|
873 |
+
"epoch": 7.565789473684211,
|
874 |
+
"grad_norm": 0.494140625,
|
875 |
+
"learning_rate": 3.3979132500809405e-05,
|
876 |
+
"loss": 0.8858,
|
877 |
+
"step": 575
|
878 |
+
},
|
879 |
+
{
|
880 |
+
"epoch": 7.631578947368421,
|
881 |
+
"grad_norm": 0.49609375,
|
882 |
+
"learning_rate": 3.227184283742591e-05,
|
883 |
+
"loss": 0.8847,
|
884 |
+
"step": 580
|
885 |
+
},
|
886 |
+
{
|
887 |
+
"epoch": 7.697368421052632,
|
888 |
+
"grad_norm": 0.48046875,
|
889 |
+
"learning_rate": 3.0600270400122335e-05,
|
890 |
+
"loss": 0.8808,
|
891 |
+
"step": 585
|
892 |
+
},
|
893 |
+
{
|
894 |
+
"epoch": 7.7631578947368425,
|
895 |
+
"grad_norm": 0.498046875,
|
896 |
+
"learning_rate": 2.89652967119336e-05,
|
897 |
+
"loss": 0.8825,
|
898 |
+
"step": 590
|
899 |
+
},
|
900 |
+
{
|
901 |
+
"epoch": 7.828947368421053,
|
902 |
+
"grad_norm": 0.470703125,
|
903 |
+
"learning_rate": 2.73677839951215e-05,
|
904 |
+
"loss": 0.8923,
|
905 |
+
"step": 595
|
906 |
+
},
|
907 |
+
{
|
908 |
+
"epoch": 7.894736842105263,
|
909 |
+
"grad_norm": 0.447265625,
|
910 |
+
"learning_rate": 2.5808574716471856e-05,
|
911 |
+
"loss": 0.8809,
|
912 |
+
"step": 600
|
913 |
+
},
|
914 |
+
{
|
915 |
+
"epoch": 7.9605263157894735,
|
916 |
+
"grad_norm": 0.490234375,
|
917 |
+
"learning_rate": 2.4288491143009795e-05,
|
918 |
+
"loss": 0.8824,
|
919 |
+
"step": 605
|
920 |
+
},
|
921 |
+
{
|
922 |
+
"epoch": 8.0,
|
923 |
+
"eval_loss": 2.547354221343994,
|
924 |
+
"eval_runtime": 0.2365,
|
925 |
+
"eval_samples_per_second": 42.286,
|
926 |
+
"eval_steps_per_second": 4.229,
|
927 |
+
"step": 608
|
928 |
+
},
|
929 |
+
{
|
930 |
+
"epoch": 8.026315789473685,
|
931 |
+
"grad_norm": 0.478515625,
|
932 |
+
"learning_rate": 2.2808334908367914e-05,
|
933 |
+
"loss": 0.8926,
|
934 |
+
"step": 610
|
935 |
+
},
|
936 |
+
{
|
937 |
+
"epoch": 8.092105263157896,
|
938 |
+
"grad_norm": 0.48828125,
|
939 |
+
"learning_rate": 2.1368886590035443e-05,
|
940 |
+
"loss": 0.8774,
|
941 |
+
"step": 615
|
942 |
+
},
|
943 |
+
{
|
944 |
+
"epoch": 8.157894736842104,
|
945 |
+
"grad_norm": 0.4921875,
|
946 |
+
"learning_rate": 1.9970905297711606e-05,
|
947 |
+
"loss": 0.8834,
|
948 |
+
"step": 620
|
949 |
+
},
|
950 |
+
{
|
951 |
+
"epoch": 8.223684210526315,
|
952 |
+
"grad_norm": 0.53125,
|
953 |
+
"learning_rate": 1.861512827298051e-05,
|
954 |
+
"loss": 0.8845,
|
955 |
+
"step": 625
|
956 |
+
},
|
957 |
+
{
|
958 |
+
"epoch": 8.289473684210526,
|
959 |
+
"grad_norm": 0.5859375,
|
960 |
+
"learning_rate": 1.7302270500518182e-05,
|
961 |
+
"loss": 0.8801,
|
962 |
+
"step": 630
|
963 |
+
},
|
964 |
+
{
|
965 |
+
"epoch": 8.355263157894736,
|
966 |
+
"grad_norm": 0.49609375,
|
967 |
+
"learning_rate": 1.6033024331037138e-05,
|
968 |
+
"loss": 0.873,
|
969 |
+
"step": 635
|
970 |
+
},
|
971 |
+
{
|
972 |
+
"epoch": 8.421052631578947,
|
973 |
+
"grad_norm": 0.5078125,
|
974 |
+
"learning_rate": 1.4808059116167305e-05,
|
975 |
+
"loss": 0.8795,
|
976 |
+
"step": 640
|
977 |
+
},
|
978 |
+
{
|
979 |
+
"epoch": 8.486842105263158,
|
980 |
+
"grad_norm": 0.466796875,
|
981 |
+
"learning_rate": 1.3628020855465572e-05,
|
982 |
+
"loss": 0.8807,
|
983 |
+
"step": 645
|
984 |
+
},
|
985 |
+
{
|
986 |
+
"epoch": 8.552631578947368,
|
987 |
+
"grad_norm": 0.46484375,
|
988 |
+
"learning_rate": 1.2493531855740625e-05,
|
989 |
+
"loss": 0.8821,
|
990 |
+
"step": 650
|
991 |
+
},
|
992 |
+
{
|
993 |
+
"epoch": 8.618421052631579,
|
994 |
+
"grad_norm": 0.47265625,
|
995 |
+
"learning_rate": 1.1405190402872202e-05,
|
996 |
+
"loss": 0.8731,
|
997 |
+
"step": 655
|
998 |
+
},
|
999 |
+
{
|
1000 |
+
"epoch": 8.68421052631579,
|
1001 |
+
"grad_norm": 0.466796875,
|
1002 |
+
"learning_rate": 1.0363570446297999e-05,
|
1003 |
+
"loss": 0.8878,
|
1004 |
+
"step": 660
|
1005 |
+
},
|
1006 |
+
{
|
1007 |
+
"epoch": 8.75,
|
1008 |
+
"grad_norm": 0.53125,
|
1009 |
+
"learning_rate": 9.369221296335006e-06,
|
1010 |
+
"loss": 0.8827,
|
1011 |
+
"step": 665
|
1012 |
+
},
|
1013 |
+
{
|
1014 |
+
"epoch": 8.81578947368421,
|
1015 |
+
"grad_norm": 0.46484375,
|
1016 |
+
"learning_rate": 8.422667334494249e-06,
|
1017 |
+
"loss": 0.8847,
|
1018 |
+
"step": 670
|
1019 |
+
},
|
1020 |
+
{
|
1021 |
+
"epoch": 8.881578947368421,
|
1022 |
+
"grad_norm": 0.48046875,
|
1023 |
+
"learning_rate": 7.524407736942174e-06,
|
1024 |
+
"loss": 0.8844,
|
1025 |
+
"step": 675
|
1026 |
+
},
|
1027 |
+
{
|
1028 |
+
"epoch": 8.947368421052632,
|
1029 |
+
"grad_norm": 0.53125,
|
1030 |
+
"learning_rate": 6.674916211254289e-06,
|
1031 |
+
"loss": 0.8759,
|
1032 |
+
"step": 680
|
1033 |
+
},
|
1034 |
+
{
|
1035 |
+
"epoch": 9.0,
|
1036 |
+
"eval_loss": 2.5541369915008545,
|
1037 |
+
"eval_runtime": 0.2345,
|
1038 |
+
"eval_samples_per_second": 42.638,
|
1039 |
+
"eval_steps_per_second": 4.264,
|
1040 |
+
"step": 684
|
1041 |
+
},
|
1042 |
+
{
|
1043 |
+
"epoch": 9.013157894736842,
|
1044 |
+
"grad_norm": 0.455078125,
|
1045 |
+
"learning_rate": 5.8746407466000464e-06,
|
1046 |
+
"loss": 0.8768,
|
1047 |
+
"step": 685
|
1048 |
+
},
|
1049 |
+
{
|
1050 |
+
"epoch": 9.078947368421053,
|
1051 |
+
"grad_norm": 0.4765625,
|
1052 |
+
"learning_rate": 5.124003377490582e-06,
|
1053 |
+
"loss": 0.8787,
|
1054 |
+
"step": 690
|
1055 |
+
},
|
1056 |
+
{
|
1057 |
+
"epoch": 9.144736842105264,
|
1058 |
+
"grad_norm": 0.515625,
|
1059 |
+
"learning_rate": 4.423399961213892e-06,
|
1060 |
+
"loss": 0.8874,
|
1061 |
+
"step": 695
|
1062 |
+
},
|
1063 |
+
{
|
1064 |
+
"epoch": 9.210526315789474,
|
1065 |
+
"grad_norm": 0.462890625,
|
1066 |
+
"learning_rate": 3.7731999690749585e-06,
|
1067 |
+
"loss": 0.875,
|
1068 |
+
"step": 700
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"epoch": 9.276315789473685,
|
1072 |
+
"grad_norm": 0.46875,
|
1073 |
+
"learning_rate": 3.1737462915508277e-06,
|
1074 |
+
"loss": 0.8777,
|
1075 |
+
"step": 705
|
1076 |
+
},
|
1077 |
+
{
|
1078 |
+
"epoch": 9.342105263157896,
|
1079 |
+
"grad_norm": 0.46875,
|
1080 |
+
"learning_rate": 2.6253550574632303e-06,
|
1081 |
+
"loss": 0.8728,
|
1082 |
+
"step": 710
|
1083 |
+
},
|
1084 |
+
{
|
1085 |
+
"epoch": 9.407894736842104,
|
1086 |
+
"grad_norm": 0.470703125,
|
1087 |
+
"learning_rate": 2.128315467264552e-06,
|
1088 |
+
"loss": 0.8791,
|
1089 |
+
"step": 715
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 9.473684210526315,
|
1093 |
+
"grad_norm": 0.4765625,
|
1094 |
+
"learning_rate": 1.6828896405244988e-06,
|
1095 |
+
"loss": 0.8809,
|
1096 |
+
"step": 720
|
1097 |
+
},
|
1098 |
+
{
|
1099 |
+
"epoch": 9.539473684210526,
|
1100 |
+
"grad_norm": 0.458984375,
|
1101 |
+
"learning_rate": 1.28931247769839e-06,
|
1102 |
+
"loss": 0.8836,
|
1103 |
+
"step": 725
|
1104 |
+
},
|
1105 |
+
{
|
1106 |
+
"epoch": 9.605263157894736,
|
1107 |
+
"grad_norm": 0.455078125,
|
1108 |
+
"learning_rate": 9.477915362496758e-07,
|
1109 |
+
"loss": 0.8819,
|
1110 |
+
"step": 730
|
1111 |
+
},
|
1112 |
+
{
|
1113 |
+
"epoch": 9.671052631578947,
|
1114 |
+
"grad_norm": 0.4609375,
|
1115 |
+
"learning_rate": 6.585069211921035e-07,
|
1116 |
+
"loss": 0.874,
|
1117 |
+
"step": 735
|
1118 |
+
},
|
1119 |
+
{
|
1120 |
+
"epoch": 9.736842105263158,
|
1121 |
+
"grad_norm": 0.458984375,
|
1122 |
+
"learning_rate": 4.216111901092501e-07,
|
1123 |
+
"loss": 0.8871,
|
1124 |
+
"step": 740
|
1125 |
+
},
|
1126 |
+
{
|
1127 |
+
"epoch": 9.802631578947368,
|
1128 |
+
"grad_norm": 0.490234375,
|
1129 |
+
"learning_rate": 2.372292727015557e-07,
|
1130 |
+
"loss": 0.8811,
|
1131 |
+
"step": 745
|
1132 |
+
},
|
1133 |
+
{
|
1134 |
+
"epoch": 9.868421052631579,
|
1135 |
+
"grad_norm": 0.45703125,
|
1136 |
+
"learning_rate": 1.0545840490313596e-07,
|
1137 |
+
"loss": 0.8768,
|
1138 |
+
"step": 750
|
1139 |
+
},
|
1140 |
+
{
|
1141 |
+
"epoch": 9.93421052631579,
|
1142 |
+
"grad_norm": 0.46875,
|
1143 |
+
"learning_rate": 2.6368077603367015e-08,
|
1144 |
+
"loss": 0.8774,
|
1145 |
+
"step": 755
|
1146 |
+
},
|
1147 |
+
{
|
1148 |
+
"epoch": 10.0,
|
1149 |
+
"grad_norm": 0.47265625,
|
1150 |
+
"learning_rate": 0.0,
|
1151 |
+
"loss": 0.8735,
|
1152 |
+
"step": 760
|
1153 |
+
},
|
1154 |
+
{
|
1155 |
+
"epoch": 10.0,
|
1156 |
+
"eval_loss": 2.5523786544799805,
|
1157 |
+
"eval_runtime": 0.2353,
|
1158 |
+
"eval_samples_per_second": 42.494,
|
1159 |
+
"eval_steps_per_second": 4.249,
|
1160 |
+
"step": 760
|
1161 |
+
},
|
1162 |
+
{
|
1163 |
+
"epoch": 10.0,
|
1164 |
+
"step": 760,
|
1165 |
+
"total_flos": 2.318334860390826e+18,
|
1166 |
+
"train_loss": 2.2621336485210217,
|
1167 |
+
"train_runtime": 1856.2046,
|
1168 |
+
"train_samples_per_second": 26.156,
|
1169 |
+
"train_steps_per_second": 0.409
|
1170 |
+
}
|
1171 |
+
],
|
1172 |
+
"logging_steps": 5,
|
1173 |
+
"max_steps": 760,
|
1174 |
+
"num_input_tokens_seen": 0,
|
1175 |
+
"num_train_epochs": 10,
|
1176 |
+
"save_steps": 100,
|
1177 |
+
"total_flos": 2.318334860390826e+18,
|
1178 |
+
"train_batch_size": 4,
|
1179 |
+
"trial_name": null,
|
1180 |
+
"trial_params": null
|
1181 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8e72d22e36be57e700734a4d42088c5a29b0893c81375cd69c80ac04f2224952
|
3 |
+
size 5176
|