Upload folder using huggingface_hub
Browse files- 49_128_e3_3e-5/README.md +63 -0
- 49_128_e3_3e-5/adapter_config.json +39 -0
- 49_128_e3_3e-5/adapter_model.safetensors +3 -0
- 49_128_e3_3e-5/all_results.json +9 -0
- 49_128_e3_3e-5/config.json +32 -0
- 49_128_e3_3e-5/merges.txt +0 -0
- 49_128_e3_3e-5/special_tokens_map.json +45 -0
- 49_128_e3_3e-5/tokenizer.json +0 -0
- 49_128_e3_3e-5/tokenizer_config.json +188 -0
- 49_128_e3_3e-5/train_results.json +9 -0
- 49_128_e3_3e-5/trainer_state.json +1604 -0
- 49_128_e3_3e-5/training_args.bin +3 -0
- 49_128_e3_3e-5/vocab.json +0 -0
49_128_e3_3e-5/README.md
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: peft
|
| 3 |
+
license: apache-2.0
|
| 4 |
+
base_model: ibm-granite/granite-3.3-8b-base
|
| 5 |
+
tags:
|
| 6 |
+
- alignment-handbook
|
| 7 |
+
- generated_from_trainer
|
| 8 |
+
datasets:
|
| 9 |
+
- data/knowledge_lora_training_data_2000
|
| 10 |
+
model-index:
|
| 11 |
+
- name: 49_128_e3_3e-5
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 16 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 17 |
+
|
| 18 |
+
# 49_128_e3_3e-5
|
| 19 |
+
|
| 20 |
+
This model is a fine-tuned version of [ibm-granite/granite-3.3-8b-base](https://huggingface.co/ibm-granite/granite-3.3-8b-base) on the data/knowledge_lora_training_data_2000 dataset.
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 3e-05
|
| 40 |
+
- train_batch_size: 2
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- distributed_type: multi-GPU
|
| 44 |
+
- num_devices: 8
|
| 45 |
+
- gradient_accumulation_steps: 2
|
| 46 |
+
- total_train_batch_size: 32
|
| 47 |
+
- total_eval_batch_size: 64
|
| 48 |
+
- optimizer: Use adamw_torch with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 49 |
+
- lr_scheduler_type: cosine
|
| 50 |
+
- lr_scheduler_warmup_ratio: 0.05
|
| 51 |
+
- num_epochs: 3.0
|
| 52 |
+
|
| 53 |
+
### Training results
|
| 54 |
+
|
| 55 |
+
|
| 56 |
+
|
| 57 |
+
### Framework versions
|
| 58 |
+
|
| 59 |
+
- PEFT 0.15.2
|
| 60 |
+
- Transformers 4.52.4
|
| 61 |
+
- Pytorch 2.7.0+cu126
|
| 62 |
+
- Datasets 3.6.0
|
| 63 |
+
- Tokenizers 0.21.2
|
49_128_e3_3e-5/adapter_config.json
ADDED
|
@@ -0,0 +1,39 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "ibm-granite/granite-3.3-8b-base",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"corda_config": null,
|
| 7 |
+
"eva_config": null,
|
| 8 |
+
"exclude_modules": null,
|
| 9 |
+
"fan_in_fan_out": false,
|
| 10 |
+
"inference_mode": true,
|
| 11 |
+
"init_lora_weights": true,
|
| 12 |
+
"layer_replication": null,
|
| 13 |
+
"layers_pattern": null,
|
| 14 |
+
"layers_to_transform": null,
|
| 15 |
+
"loftq_config": {},
|
| 16 |
+
"lora_alpha": 256,
|
| 17 |
+
"lora_bias": false,
|
| 18 |
+
"lora_dropout": 0.05,
|
| 19 |
+
"megatron_config": null,
|
| 20 |
+
"megatron_core": "megatron.core",
|
| 21 |
+
"modules_to_save": null,
|
| 22 |
+
"peft_type": "LORA",
|
| 23 |
+
"r": 128,
|
| 24 |
+
"rank_pattern": {},
|
| 25 |
+
"revision": null,
|
| 26 |
+
"target_modules": [
|
| 27 |
+
"down_proj",
|
| 28 |
+
"q_proj",
|
| 29 |
+
"gate_proj",
|
| 30 |
+
"k_proj",
|
| 31 |
+
"o_proj",
|
| 32 |
+
"up_proj",
|
| 33 |
+
"v_proj"
|
| 34 |
+
],
|
| 35 |
+
"task_type": "CAUSAL_LM",
|
| 36 |
+
"trainable_token_indices": null,
|
| 37 |
+
"use_dora": false,
|
| 38 |
+
"use_rslora": false
|
| 39 |
+
}
|
49_128_e3_3e-5/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0a739c1de2b7a8cbbf00ddeb310b7588498e5327213360541f505523f92c69ce
|
| 3 |
+
size 791751704
|
49_128_e3_3e-5/all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.5875073592094884e+18,
|
| 4 |
+
"train_loss": 0.5184633527551928,
|
| 5 |
+
"train_runtime": 742.8841,
|
| 6 |
+
"train_samples": 11877,
|
| 7 |
+
"train_samples_per_second": 47.963,
|
| 8 |
+
"train_steps_per_second": 1.502
|
| 9 |
+
}
|
49_128_e3_3e-5/config.json
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"GraniteForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0,
|
| 7 |
+
"attention_multiplier": 0.0078125,
|
| 8 |
+
"bos_token_id": 0,
|
| 9 |
+
"embedding_multiplier": 12.0,
|
| 10 |
+
"eos_token_id": 0,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 4096,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 12800,
|
| 15 |
+
"logits_scaling": 16.0,
|
| 16 |
+
"max_position_embeddings": 131072,
|
| 17 |
+
"mlp_bias": false,
|
| 18 |
+
"model_type": "granite",
|
| 19 |
+
"num_attention_heads": 32,
|
| 20 |
+
"num_hidden_layers": 40,
|
| 21 |
+
"num_key_value_heads": 8,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"residual_multiplier": 0.22,
|
| 24 |
+
"rms_norm_eps": 1e-05,
|
| 25 |
+
"rope_scaling": null,
|
| 26 |
+
"rope_theta": 10000000.0,
|
| 27 |
+
"tie_word_embeddings": true,
|
| 28 |
+
"torch_dtype": "bfloat16",
|
| 29 |
+
"transformers_version": "4.52.4",
|
| 30 |
+
"use_cache": true,
|
| 31 |
+
"vocab_size": 49152
|
| 32 |
+
}
|
49_128_e3_3e-5/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
49_128_e3_3e-5/special_tokens_map.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|endoftext|>",
|
| 4 |
+
"<fim_prefix>",
|
| 5 |
+
"<fim_middle>",
|
| 6 |
+
"<fim_suffix>",
|
| 7 |
+
"<fim_pad>",
|
| 8 |
+
"<filename>",
|
| 9 |
+
"<gh_stars>",
|
| 10 |
+
"<issue_start>",
|
| 11 |
+
"<issue_comment>",
|
| 12 |
+
"<issue_closed>",
|
| 13 |
+
"<jupyter_start>",
|
| 14 |
+
"<jupyter_text>",
|
| 15 |
+
"<jupyter_code>",
|
| 16 |
+
"<jupyter_output>",
|
| 17 |
+
"<empty_output>",
|
| 18 |
+
"<commit_before>",
|
| 19 |
+
"<commit_msg>",
|
| 20 |
+
"<commit_after>",
|
| 21 |
+
"<reponame>"
|
| 22 |
+
],
|
| 23 |
+
"bos_token": {
|
| 24 |
+
"content": "<|endoftext|>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
},
|
| 30 |
+
"eos_token": {
|
| 31 |
+
"content": "<|endoftext|>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false
|
| 36 |
+
},
|
| 37 |
+
"pad_token": "<reponame>",
|
| 38 |
+
"unk_token": {
|
| 39 |
+
"content": "<|endoftext|>",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": false,
|
| 42 |
+
"rstrip": false,
|
| 43 |
+
"single_word": false
|
| 44 |
+
}
|
| 45 |
+
}
|
49_128_e3_3e-5/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
49_128_e3_3e-5/tokenizer_config.json
ADDED
|
@@ -0,0 +1,188 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"0": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"1": {
|
| 13 |
+
"content": "<fim_prefix>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"2": {
|
| 21 |
+
"content": "<fim_middle>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"3": {
|
| 29 |
+
"content": "<fim_suffix>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
},
|
| 36 |
+
"4": {
|
| 37 |
+
"content": "<fim_pad>",
|
| 38 |
+
"lstrip": false,
|
| 39 |
+
"normalized": false,
|
| 40 |
+
"rstrip": false,
|
| 41 |
+
"single_word": false,
|
| 42 |
+
"special": true
|
| 43 |
+
},
|
| 44 |
+
"5": {
|
| 45 |
+
"content": "<filename>",
|
| 46 |
+
"lstrip": false,
|
| 47 |
+
"normalized": false,
|
| 48 |
+
"rstrip": false,
|
| 49 |
+
"single_word": false,
|
| 50 |
+
"special": true
|
| 51 |
+
},
|
| 52 |
+
"6": {
|
| 53 |
+
"content": "<gh_stars>",
|
| 54 |
+
"lstrip": false,
|
| 55 |
+
"normalized": false,
|
| 56 |
+
"rstrip": false,
|
| 57 |
+
"single_word": false,
|
| 58 |
+
"special": true
|
| 59 |
+
},
|
| 60 |
+
"7": {
|
| 61 |
+
"content": "<issue_start>",
|
| 62 |
+
"lstrip": false,
|
| 63 |
+
"normalized": false,
|
| 64 |
+
"rstrip": false,
|
| 65 |
+
"single_word": false,
|
| 66 |
+
"special": true
|
| 67 |
+
},
|
| 68 |
+
"8": {
|
| 69 |
+
"content": "<issue_comment>",
|
| 70 |
+
"lstrip": false,
|
| 71 |
+
"normalized": false,
|
| 72 |
+
"rstrip": false,
|
| 73 |
+
"single_word": false,
|
| 74 |
+
"special": true
|
| 75 |
+
},
|
| 76 |
+
"9": {
|
| 77 |
+
"content": "<issue_closed>",
|
| 78 |
+
"lstrip": false,
|
| 79 |
+
"normalized": false,
|
| 80 |
+
"rstrip": false,
|
| 81 |
+
"single_word": false,
|
| 82 |
+
"special": true
|
| 83 |
+
},
|
| 84 |
+
"10": {
|
| 85 |
+
"content": "<jupyter_start>",
|
| 86 |
+
"lstrip": false,
|
| 87 |
+
"normalized": false,
|
| 88 |
+
"rstrip": false,
|
| 89 |
+
"single_word": false,
|
| 90 |
+
"special": true
|
| 91 |
+
},
|
| 92 |
+
"11": {
|
| 93 |
+
"content": "<jupyter_text>",
|
| 94 |
+
"lstrip": false,
|
| 95 |
+
"normalized": false,
|
| 96 |
+
"rstrip": false,
|
| 97 |
+
"single_word": false,
|
| 98 |
+
"special": true
|
| 99 |
+
},
|
| 100 |
+
"12": {
|
| 101 |
+
"content": "<jupyter_code>",
|
| 102 |
+
"lstrip": false,
|
| 103 |
+
"normalized": false,
|
| 104 |
+
"rstrip": false,
|
| 105 |
+
"single_word": false,
|
| 106 |
+
"special": true
|
| 107 |
+
},
|
| 108 |
+
"13": {
|
| 109 |
+
"content": "<jupyter_output>",
|
| 110 |
+
"lstrip": false,
|
| 111 |
+
"normalized": false,
|
| 112 |
+
"rstrip": false,
|
| 113 |
+
"single_word": false,
|
| 114 |
+
"special": true
|
| 115 |
+
},
|
| 116 |
+
"14": {
|
| 117 |
+
"content": "<empty_output>",
|
| 118 |
+
"lstrip": false,
|
| 119 |
+
"normalized": false,
|
| 120 |
+
"rstrip": false,
|
| 121 |
+
"single_word": false,
|
| 122 |
+
"special": true
|
| 123 |
+
},
|
| 124 |
+
"15": {
|
| 125 |
+
"content": "<commit_before>",
|
| 126 |
+
"lstrip": false,
|
| 127 |
+
"normalized": false,
|
| 128 |
+
"rstrip": false,
|
| 129 |
+
"single_word": false,
|
| 130 |
+
"special": true
|
| 131 |
+
},
|
| 132 |
+
"16": {
|
| 133 |
+
"content": "<commit_msg>",
|
| 134 |
+
"lstrip": false,
|
| 135 |
+
"normalized": false,
|
| 136 |
+
"rstrip": false,
|
| 137 |
+
"single_word": false,
|
| 138 |
+
"special": true
|
| 139 |
+
},
|
| 140 |
+
"17": {
|
| 141 |
+
"content": "<commit_after>",
|
| 142 |
+
"lstrip": false,
|
| 143 |
+
"normalized": false,
|
| 144 |
+
"rstrip": false,
|
| 145 |
+
"single_word": false,
|
| 146 |
+
"special": true
|
| 147 |
+
},
|
| 148 |
+
"18": {
|
| 149 |
+
"content": "<reponame>",
|
| 150 |
+
"lstrip": false,
|
| 151 |
+
"normalized": false,
|
| 152 |
+
"rstrip": false,
|
| 153 |
+
"single_word": false,
|
| 154 |
+
"special": true
|
| 155 |
+
}
|
| 156 |
+
},
|
| 157 |
+
"additional_special_tokens": [
|
| 158 |
+
"<|endoftext|>",
|
| 159 |
+
"<fim_prefix>",
|
| 160 |
+
"<fim_middle>",
|
| 161 |
+
"<fim_suffix>",
|
| 162 |
+
"<fim_pad>",
|
| 163 |
+
"<filename>",
|
| 164 |
+
"<gh_stars>",
|
| 165 |
+
"<issue_start>",
|
| 166 |
+
"<issue_comment>",
|
| 167 |
+
"<issue_closed>",
|
| 168 |
+
"<jupyter_start>",
|
| 169 |
+
"<jupyter_text>",
|
| 170 |
+
"<jupyter_code>",
|
| 171 |
+
"<jupyter_output>",
|
| 172 |
+
"<empty_output>",
|
| 173 |
+
"<commit_before>",
|
| 174 |
+
"<commit_msg>",
|
| 175 |
+
"<commit_after>",
|
| 176 |
+
"<reponame>"
|
| 177 |
+
],
|
| 178 |
+
"bos_token": "<|endoftext|>",
|
| 179 |
+
"clean_up_tokenization_spaces": true,
|
| 180 |
+
"eos_token": "<|endoftext|>",
|
| 181 |
+
"extra_special_tokens": {},
|
| 182 |
+
"model_max_length": 8192,
|
| 183 |
+
"pad_token": "<reponame>",
|
| 184 |
+
"padding_side": "left",
|
| 185 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 186 |
+
"unk_token": "<|endoftext|>",
|
| 187 |
+
"vocab_size": 49152
|
| 188 |
+
}
|
49_128_e3_3e-5/train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.5875073592094884e+18,
|
| 4 |
+
"train_loss": 0.5184633527551928,
|
| 5 |
+
"train_runtime": 742.8841,
|
| 6 |
+
"train_samples": 11877,
|
| 7 |
+
"train_samples_per_second": 47.963,
|
| 8 |
+
"train_steps_per_second": 1.502
|
| 9 |
+
}
|
49_128_e3_3e-5/trainer_state.json
ADDED
|
@@ -0,0 +1,1604 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1116,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.013458950201884253,
|
| 14 |
+
"grad_norm": 0.9882215261459351,
|
| 15 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 16 |
+
"loss": 1.2639,
|
| 17 |
+
"step": 5
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.026917900403768506,
|
| 21 |
+
"grad_norm": 0.940019965171814,
|
| 22 |
+
"learning_rate": 4.821428571428572e-06,
|
| 23 |
+
"loss": 1.2334,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.040376850605652756,
|
| 28 |
+
"grad_norm": 0.6706891059875488,
|
| 29 |
+
"learning_rate": 7.5e-06,
|
| 30 |
+
"loss": 1.2291,
|
| 31 |
+
"step": 15
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.05383580080753701,
|
| 35 |
+
"grad_norm": 0.7410071492195129,
|
| 36 |
+
"learning_rate": 1.0178571428571429e-05,
|
| 37 |
+
"loss": 1.2068,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.06729475100942127,
|
| 42 |
+
"grad_norm": 1.317247748374939,
|
| 43 |
+
"learning_rate": 1.2857142857142857e-05,
|
| 44 |
+
"loss": 1.209,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.08075370121130551,
|
| 49 |
+
"grad_norm": 0.6265580654144287,
|
| 50 |
+
"learning_rate": 1.553571428571429e-05,
|
| 51 |
+
"loss": 1.1986,
|
| 52 |
+
"step": 30
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.09421265141318977,
|
| 56 |
+
"grad_norm": 0.6021242141723633,
|
| 57 |
+
"learning_rate": 1.8214285714285712e-05,
|
| 58 |
+
"loss": 1.1809,
|
| 59 |
+
"step": 35
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.10767160161507403,
|
| 63 |
+
"grad_norm": 0.5548004508018494,
|
| 64 |
+
"learning_rate": 2.089285714285714e-05,
|
| 65 |
+
"loss": 1.1811,
|
| 66 |
+
"step": 40
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.12113055181695828,
|
| 70 |
+
"grad_norm": 0.5183567404747009,
|
| 71 |
+
"learning_rate": 2.357142857142857e-05,
|
| 72 |
+
"loss": 1.1822,
|
| 73 |
+
"step": 45
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.13458950201884254,
|
| 77 |
+
"grad_norm": 0.5572848320007324,
|
| 78 |
+
"learning_rate": 2.625e-05,
|
| 79 |
+
"loss": 1.1761,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.1480484522207268,
|
| 84 |
+
"grad_norm": 0.5353060364723206,
|
| 85 |
+
"learning_rate": 2.892857142857143e-05,
|
| 86 |
+
"loss": 1.124,
|
| 87 |
+
"step": 55
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.16150740242261102,
|
| 91 |
+
"grad_norm": 0.6251804828643799,
|
| 92 |
+
"learning_rate": 2.9999407089793328e-05,
|
| 93 |
+
"loss": 1.1083,
|
| 94 |
+
"step": 60
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.17496635262449528,
|
| 98 |
+
"grad_norm": 0.6196370124816895,
|
| 99 |
+
"learning_rate": 2.99957839193835e-05,
|
| 100 |
+
"loss": 1.0978,
|
| 101 |
+
"step": 65
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.18842530282637954,
|
| 105 |
+
"grad_norm": 0.4450572431087494,
|
| 106 |
+
"learning_rate": 2.9988867767784412e-05,
|
| 107 |
+
"loss": 1.0598,
|
| 108 |
+
"step": 70
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.2018842530282638,
|
| 112 |
+
"grad_norm": 0.6395757794380188,
|
| 113 |
+
"learning_rate": 2.997866015374007e-05,
|
| 114 |
+
"loss": 1.125,
|
| 115 |
+
"step": 75
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.21534320323014805,
|
| 119 |
+
"grad_norm": 0.5264338254928589,
|
| 120 |
+
"learning_rate": 2.996516331877925e-05,
|
| 121 |
+
"loss": 1.0484,
|
| 122 |
+
"step": 80
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2288021534320323,
|
| 126 |
+
"grad_norm": 0.5416696667671204,
|
| 127 |
+
"learning_rate": 2.994838022672324e-05,
|
| 128 |
+
"loss": 1.0364,
|
| 129 |
+
"step": 85
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.24226110363391656,
|
| 133 |
+
"grad_norm": 0.4816514253616333,
|
| 134 |
+
"learning_rate": 2.9928314563035015e-05,
|
| 135 |
+
"loss": 1.0884,
|
| 136 |
+
"step": 90
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.2557200538358008,
|
| 140 |
+
"grad_norm": 0.8453904986381531,
|
| 141 |
+
"learning_rate": 2.9904970734009933e-05,
|
| 142 |
+
"loss": 1.0505,
|
| 143 |
+
"step": 95
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.2691790040376851,
|
| 147 |
+
"grad_norm": 0.5211928486824036,
|
| 148 |
+
"learning_rate": 2.9878353865808144e-05,
|
| 149 |
+
"loss": 0.9872,
|
| 150 |
+
"step": 100
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.28263795423956933,
|
| 154 |
+
"grad_norm": 0.5633994936943054,
|
| 155 |
+
"learning_rate": 2.9848469803328896e-05,
|
| 156 |
+
"loss": 1.0204,
|
| 157 |
+
"step": 105
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.2960969044414536,
|
| 161 |
+
"grad_norm": 0.5428367853164673,
|
| 162 |
+
"learning_rate": 2.981532510892707e-05,
|
| 163 |
+
"loss": 0.9901,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.30955585464333785,
|
| 168 |
+
"grad_norm": 0.5815945863723755,
|
| 169 |
+
"learning_rate": 2.9778927060972075e-05,
|
| 170 |
+
"loss": 1.0007,
|
| 171 |
+
"step": 115
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.32301480484522205,
|
| 175 |
+
"grad_norm": 0.5644654035568237,
|
| 176 |
+
"learning_rate": 2.9739283652249625e-05,
|
| 177 |
+
"loss": 0.9891,
|
| 178 |
+
"step": 120
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3364737550471063,
|
| 182 |
+
"grad_norm": 0.633641242980957,
|
| 183 |
+
"learning_rate": 2.9696403588206517e-05,
|
| 184 |
+
"loss": 0.9361,
|
| 185 |
+
"step": 125
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.34993270524899056,
|
| 189 |
+
"grad_norm": 0.6438229084014893,
|
| 190 |
+
"learning_rate": 2.9650296285038996e-05,
|
| 191 |
+
"loss": 0.9631,
|
| 192 |
+
"step": 130
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.3633916554508748,
|
| 196 |
+
"grad_norm": 0.6100443601608276,
|
| 197 |
+
"learning_rate": 2.9600971867625027e-05,
|
| 198 |
+
"loss": 0.9458,
|
| 199 |
+
"step": 135
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.3768506056527591,
|
| 203 |
+
"grad_norm": 0.6567280292510986,
|
| 204 |
+
"learning_rate": 2.95484411673009e-05,
|
| 205 |
+
"loss": 0.9037,
|
| 206 |
+
"step": 140
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.39030955585464333,
|
| 210 |
+
"grad_norm": 0.6098927855491638,
|
| 211 |
+
"learning_rate": 2.9492715719482776e-05,
|
| 212 |
+
"loss": 0.9394,
|
| 213 |
+
"step": 145
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.4037685060565276,
|
| 217 |
+
"grad_norm": 0.6474848389625549,
|
| 218 |
+
"learning_rate": 2.9433807761133542e-05,
|
| 219 |
+
"loss": 0.9386,
|
| 220 |
+
"step": 150
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.41722745625841184,
|
| 224 |
+
"grad_norm": 0.7610502243041992,
|
| 225 |
+
"learning_rate": 2.9371730228075684e-05,
|
| 226 |
+
"loss": 0.8807,
|
| 227 |
+
"step": 155
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4306864064602961,
|
| 231 |
+
"grad_norm": 0.7775782346725464,
|
| 232 |
+
"learning_rate": 2.9306496752150638e-05,
|
| 233 |
+
"loss": 0.8372,
|
| 234 |
+
"step": 160
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.44414535666218036,
|
| 238 |
+
"grad_norm": 0.7716594934463501,
|
| 239 |
+
"learning_rate": 2.9238121658225337e-05,
|
| 240 |
+
"loss": 0.8837,
|
| 241 |
+
"step": 165
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.4576043068640646,
|
| 245 |
+
"grad_norm": 0.6561233401298523,
|
| 246 |
+
"learning_rate": 2.9166619961046545e-05,
|
| 247 |
+
"loss": 0.8971,
|
| 248 |
+
"step": 170
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.47106325706594887,
|
| 252 |
+
"grad_norm": 0.698351263999939,
|
| 253 |
+
"learning_rate": 2.909200736194372e-05,
|
| 254 |
+
"loss": 0.8761,
|
| 255 |
+
"step": 175
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.4845222072678331,
|
| 259 |
+
"grad_norm": 0.6881304383277893,
|
| 260 |
+
"learning_rate": 2.9014300245381095e-05,
|
| 261 |
+
"loss": 0.834,
|
| 262 |
+
"step": 180
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4979811574697174,
|
| 266 |
+
"grad_norm": 0.726060152053833,
|
| 267 |
+
"learning_rate": 2.8933515675359766e-05,
|
| 268 |
+
"loss": 0.9004,
|
| 269 |
+
"step": 185
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.5114401076716016,
|
| 273 |
+
"grad_norm": 0.649829626083374,
|
| 274 |
+
"learning_rate": 2.8849671391670518e-05,
|
| 275 |
+
"loss": 0.837,
|
| 276 |
+
"step": 190
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.5248990578734859,
|
| 280 |
+
"grad_norm": 0.7975636124610901,
|
| 281 |
+
"learning_rate": 2.8762785805998295e-05,
|
| 282 |
+
"loss": 0.8313,
|
| 283 |
+
"step": 195
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.5383580080753702,
|
| 287 |
+
"grad_norm": 0.855278730392456,
|
| 288 |
+
"learning_rate": 2.86728779978791e-05,
|
| 289 |
+
"loss": 0.8254,
|
| 290 |
+
"step": 200
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.5518169582772544,
|
| 294 |
+
"grad_norm": 0.829576313495636,
|
| 295 |
+
"learning_rate": 2.8579967710510264e-05,
|
| 296 |
+
"loss": 0.8876,
|
| 297 |
+
"step": 205
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.5652759084791387,
|
| 301 |
+
"grad_norm": 0.7353431582450867,
|
| 302 |
+
"learning_rate": 2.8484075346414936e-05,
|
| 303 |
+
"loss": 0.7699,
|
| 304 |
+
"step": 210
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5787348586810229,
|
| 308 |
+
"grad_norm": 0.8122780323028564,
|
| 309 |
+
"learning_rate": 2.838522196296182e-05,
|
| 310 |
+
"loss": 0.8452,
|
| 311 |
+
"step": 215
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.5921938088829072,
|
| 315 |
+
"grad_norm": 0.7391046285629272,
|
| 316 |
+
"learning_rate": 2.8283429267741138e-05,
|
| 317 |
+
"loss": 0.7429,
|
| 318 |
+
"step": 220
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6056527590847914,
|
| 322 |
+
"grad_norm": 1.0399044752120972,
|
| 323 |
+
"learning_rate": 2.8178719613797747e-05,
|
| 324 |
+
"loss": 0.7926,
|
| 325 |
+
"step": 225
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.6191117092866757,
|
| 329 |
+
"grad_norm": 0.8288000226020813,
|
| 330 |
+
"learning_rate": 2.807111599472254e-05,
|
| 331 |
+
"loss": 0.765,
|
| 332 |
+
"step": 230
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.6325706594885598,
|
| 336 |
+
"grad_norm": 0.9703077673912048,
|
| 337 |
+
"learning_rate": 2.7960642039603235e-05,
|
| 338 |
+
"loss": 0.7335,
|
| 339 |
+
"step": 235
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.6460296096904441,
|
| 343 |
+
"grad_norm": 0.9160774946212769,
|
| 344 |
+
"learning_rate": 2.7847322007835546e-05,
|
| 345 |
+
"loss": 0.782,
|
| 346 |
+
"step": 240
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.6594885598923284,
|
| 350 |
+
"grad_norm": 0.9156641960144043,
|
| 351 |
+
"learning_rate": 2.773118078379597e-05,
|
| 352 |
+
"loss": 0.7643,
|
| 353 |
+
"step": 245
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.6729475100942126,
|
| 357 |
+
"grad_norm": 0.8600050210952759,
|
| 358 |
+
"learning_rate": 2.7612243871377342e-05,
|
| 359 |
+
"loss": 0.7607,
|
| 360 |
+
"step": 250
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.6864064602960969,
|
| 364 |
+
"grad_norm": 0.7924662828445435,
|
| 365 |
+
"learning_rate": 2.749053738838834e-05,
|
| 366 |
+
"loss": 0.7422,
|
| 367 |
+
"step": 255
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.6998654104979811,
|
| 371 |
+
"grad_norm": 0.799248993396759,
|
| 372 |
+
"learning_rate": 2.7366088060818154e-05,
|
| 373 |
+
"loss": 0.7405,
|
| 374 |
+
"step": 260
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.7133243606998654,
|
| 378 |
+
"grad_norm": 0.7867820858955383,
|
| 379 |
+
"learning_rate": 2.7238923216967666e-05,
|
| 380 |
+
"loss": 0.7193,
|
| 381 |
+
"step": 265
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.7267833109017496,
|
| 385 |
+
"grad_norm": 0.8833760619163513,
|
| 386 |
+
"learning_rate": 2.7109070781448283e-05,
|
| 387 |
+
"loss": 0.7143,
|
| 388 |
+
"step": 270
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.7402422611036339,
|
| 392 |
+
"grad_norm": 0.9755755066871643,
|
| 393 |
+
"learning_rate": 2.6976559269049875e-05,
|
| 394 |
+
"loss": 0.737,
|
| 395 |
+
"step": 275
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.7537012113055181,
|
| 399 |
+
"grad_norm": 0.9140403270721436,
|
| 400 |
+
"learning_rate": 2.6841417778479132e-05,
|
| 401 |
+
"loss": 0.6613,
|
| 402 |
+
"step": 280
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.7671601615074024,
|
| 406 |
+
"grad_norm": 1.0706220865249634,
|
| 407 |
+
"learning_rate": 2.670367598596963e-05,
|
| 408 |
+
"loss": 0.6931,
|
| 409 |
+
"step": 285
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.7806191117092867,
|
| 413 |
+
"grad_norm": 0.895652174949646,
|
| 414 |
+
"learning_rate": 2.6563364138765137e-05,
|
| 415 |
+
"loss": 0.6678,
|
| 416 |
+
"step": 290
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7940780619111709,
|
| 420 |
+
"grad_norm": 0.9422807097434998,
|
| 421 |
+
"learning_rate": 2.6420513048477503e-05,
|
| 422 |
+
"loss": 0.7114,
|
| 423 |
+
"step": 295
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.8075370121130552,
|
| 427 |
+
"grad_norm": 0.9548665285110474,
|
| 428 |
+
"learning_rate": 2.6275154084320622e-05,
|
| 429 |
+
"loss": 0.6963,
|
| 430 |
+
"step": 300
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.8209959623149394,
|
| 434 |
+
"grad_norm": 0.8600846529006958,
|
| 435 |
+
"learning_rate": 2.6127319166221923e-05,
|
| 436 |
+
"loss": 0.6485,
|
| 437 |
+
"step": 305
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.8344549125168237,
|
| 441 |
+
"grad_norm": 0.8546967506408691,
|
| 442 |
+
"learning_rate": 2.5977040757812997e-05,
|
| 443 |
+
"loss": 0.6661,
|
| 444 |
+
"step": 310
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.847913862718708,
|
| 448 |
+
"grad_norm": 0.9586823582649231,
|
| 449 |
+
"learning_rate": 2.5824351859300748e-05,
|
| 450 |
+
"loss": 0.6856,
|
| 451 |
+
"step": 315
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.8613728129205922,
|
| 455 |
+
"grad_norm": 1.010869026184082,
|
| 456 |
+
"learning_rate": 2.5669286000220757e-05,
|
| 457 |
+
"loss": 0.701,
|
| 458 |
+
"step": 320
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.8748317631224765,
|
| 462 |
+
"grad_norm": 1.1527146100997925,
|
| 463 |
+
"learning_rate": 2.551187723207443e-05,
|
| 464 |
+
"loss": 0.6431,
|
| 465 |
+
"step": 325
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.8882907133243607,
|
| 469 |
+
"grad_norm": 0.948448657989502,
|
| 470 |
+
"learning_rate": 2.5352160120851464e-05,
|
| 471 |
+
"loss": 0.6379,
|
| 472 |
+
"step": 330
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.901749663526245,
|
| 476 |
+
"grad_norm": 1.0279428958892822,
|
| 477 |
+
"learning_rate": 2.519016973943939e-05,
|
| 478 |
+
"loss": 0.6278,
|
| 479 |
+
"step": 335
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.9152086137281292,
|
| 483 |
+
"grad_norm": 0.938048243522644,
|
| 484 |
+
"learning_rate": 2.5025941659921783e-05,
|
| 485 |
+
"loss": 0.6158,
|
| 486 |
+
"step": 340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.9286675639300135,
|
| 490 |
+
"grad_norm": 0.951744794845581,
|
| 491 |
+
"learning_rate": 2.485951194576685e-05,
|
| 492 |
+
"loss": 0.6429,
|
| 493 |
+
"step": 345
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.9421265141318977,
|
| 497 |
+
"grad_norm": 0.9467681646347046,
|
| 498 |
+
"learning_rate": 2.469091714390811e-05,
|
| 499 |
+
"loss": 0.6459,
|
| 500 |
+
"step": 350
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.955585464333782,
|
| 504 |
+
"grad_norm": 1.0841658115386963,
|
| 505 |
+
"learning_rate": 2.4520194276718938e-05,
|
| 506 |
+
"loss": 0.6094,
|
| 507 |
+
"step": 355
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.9690444145356663,
|
| 511 |
+
"grad_norm": 0.8894978165626526,
|
| 512 |
+
"learning_rate": 2.4347380833882634e-05,
|
| 513 |
+
"loss": 0.623,
|
| 514 |
+
"step": 360
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.9825033647375505,
|
| 518 |
+
"grad_norm": 0.9439630508422852,
|
| 519 |
+
"learning_rate": 2.417251476415998e-05,
|
| 520 |
+
"loss": 0.6406,
|
| 521 |
+
"step": 365
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.9959623149394348,
|
| 525 |
+
"grad_norm": 0.9258154630661011,
|
| 526 |
+
"learning_rate": 2.3995634467055882e-05,
|
| 527 |
+
"loss": 0.5976,
|
| 528 |
+
"step": 370
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.0080753701211305,
|
| 532 |
+
"grad_norm": 1.1117361783981323,
|
| 533 |
+
"learning_rate": 2.3816778784387097e-05,
|
| 534 |
+
"loss": 0.5804,
|
| 535 |
+
"step": 375
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.0215343203230147,
|
| 539 |
+
"grad_norm": 0.9733107089996338,
|
| 540 |
+
"learning_rate": 2.363598699175281e-05,
|
| 541 |
+
"loss": 0.5061,
|
| 542 |
+
"step": 380
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.034993270524899,
|
| 546 |
+
"grad_norm": 0.9482859969139099,
|
| 547 |
+
"learning_rate": 2.3453298789909935e-05,
|
| 548 |
+
"loss": 0.523,
|
| 549 |
+
"step": 385
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.0484522207267832,
|
| 553 |
+
"grad_norm": 1.136534333229065,
|
| 554 |
+
"learning_rate": 2.3268754296055122e-05,
|
| 555 |
+
"loss": 0.5518,
|
| 556 |
+
"step": 390
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.0619111709286675,
|
| 560 |
+
"grad_norm": 1.026702880859375,
|
| 561 |
+
"learning_rate": 2.3082394035015212e-05,
|
| 562 |
+
"loss": 0.5276,
|
| 563 |
+
"step": 395
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.0753701211305517,
|
| 567 |
+
"grad_norm": 1.0939162969589233,
|
| 568 |
+
"learning_rate": 2.2894258930348284e-05,
|
| 569 |
+
"loss": 0.5324,
|
| 570 |
+
"step": 400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.088829071332436,
|
| 574 |
+
"grad_norm": 0.9696338176727295,
|
| 575 |
+
"learning_rate": 2.2704390295357056e-05,
|
| 576 |
+
"loss": 0.5136,
|
| 577 |
+
"step": 405
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.1022880215343203,
|
| 581 |
+
"grad_norm": 0.9916676878929138,
|
| 582 |
+
"learning_rate": 2.2512829824016773e-05,
|
| 583 |
+
"loss": 0.5235,
|
| 584 |
+
"step": 410
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.1157469717362045,
|
| 588 |
+
"grad_norm": 1.096625804901123,
|
| 589 |
+
"learning_rate": 2.2319619581819458e-05,
|
| 590 |
+
"loss": 0.4873,
|
| 591 |
+
"step": 415
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.1292059219380888,
|
| 595 |
+
"grad_norm": 1.1451526880264282,
|
| 596 |
+
"learning_rate": 2.2124801996536575e-05,
|
| 597 |
+
"loss": 0.4765,
|
| 598 |
+
"step": 420
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.142664872139973,
|
| 602 |
+
"grad_norm": 0.9332502484321594,
|
| 603 |
+
"learning_rate": 2.1928419848902163e-05,
|
| 604 |
+
"loss": 0.4786,
|
| 605 |
+
"step": 425
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.1561238223418573,
|
| 609 |
+
"grad_norm": 1.0080773830413818,
|
| 610 |
+
"learning_rate": 2.1730516263218472e-05,
|
| 611 |
+
"loss": 0.4909,
|
| 612 |
+
"step": 430
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.1695827725437415,
|
| 616 |
+
"grad_norm": 1.092008113861084,
|
| 617 |
+
"learning_rate": 2.1531134697886103e-05,
|
| 618 |
+
"loss": 0.4696,
|
| 619 |
+
"step": 435
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.1830417227456258,
|
| 623 |
+
"grad_norm": 1.1402854919433594,
|
| 624 |
+
"learning_rate": 2.133031893586083e-05,
|
| 625 |
+
"loss": 0.4854,
|
| 626 |
+
"step": 440
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.19650067294751,
|
| 630 |
+
"grad_norm": 1.10186767578125,
|
| 631 |
+
"learning_rate": 2.1128113075039124e-05,
|
| 632 |
+
"loss": 0.4958,
|
| 633 |
+
"step": 445
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.2099596231493943,
|
| 637 |
+
"grad_norm": 1.1457712650299072,
|
| 638 |
+
"learning_rate": 2.092456151857455e-05,
|
| 639 |
+
"loss": 0.4801,
|
| 640 |
+
"step": 450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.2234185733512786,
|
| 644 |
+
"grad_norm": 0.9913672804832458,
|
| 645 |
+
"learning_rate": 2.0719708965127073e-05,
|
| 646 |
+
"loss": 0.5022,
|
| 647 |
+
"step": 455
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.2368775235531628,
|
| 651 |
+
"grad_norm": 1.1649874448776245,
|
| 652 |
+
"learning_rate": 2.0513600399047545e-05,
|
| 653 |
+
"loss": 0.4768,
|
| 654 |
+
"step": 460
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.250336473755047,
|
| 658 |
+
"grad_norm": 0.9590378999710083,
|
| 659 |
+
"learning_rate": 2.0306281080499413e-05,
|
| 660 |
+
"loss": 0.5171,
|
| 661 |
+
"step": 465
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.2637954239569313,
|
| 665 |
+
"grad_norm": 1.1455888748168945,
|
| 666 |
+
"learning_rate": 2.009779653551983e-05,
|
| 667 |
+
"loss": 0.4673,
|
| 668 |
+
"step": 470
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.2772543741588156,
|
| 672 |
+
"grad_norm": 1.0884604454040527,
|
| 673 |
+
"learning_rate": 1.9888192546022488e-05,
|
| 674 |
+
"loss": 0.4416,
|
| 675 |
+
"step": 475
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.2907133243606999,
|
| 679 |
+
"grad_norm": 1.0540392398834229,
|
| 680 |
+
"learning_rate": 1.9677515139744126e-05,
|
| 681 |
+
"loss": 0.5134,
|
| 682 |
+
"step": 480
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.304172274562584,
|
| 686 |
+
"grad_norm": 0.9360066056251526,
|
| 687 |
+
"learning_rate": 1.946581058013717e-05,
|
| 688 |
+
"loss": 0.4749,
|
| 689 |
+
"step": 485
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.3176312247644684,
|
| 693 |
+
"grad_norm": 1.0760797262191772,
|
| 694 |
+
"learning_rate": 1.9253125356210547e-05,
|
| 695 |
+
"loss": 0.461,
|
| 696 |
+
"step": 490
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.3310901749663526,
|
| 700 |
+
"grad_norm": 1.1534333229064941,
|
| 701 |
+
"learning_rate": 1.903950617232098e-05,
|
| 702 |
+
"loss": 0.4503,
|
| 703 |
+
"step": 495
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.3445491251682369,
|
| 707 |
+
"grad_norm": 1.110041618347168,
|
| 708 |
+
"learning_rate": 1.8824999937917025e-05,
|
| 709 |
+
"loss": 0.4331,
|
| 710 |
+
"step": 500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.3580080753701211,
|
| 714 |
+
"grad_norm": 1.0444817543029785,
|
| 715 |
+
"learning_rate": 1.8609653757238026e-05,
|
| 716 |
+
"loss": 0.4292,
|
| 717 |
+
"step": 505
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.3714670255720054,
|
| 721 |
+
"grad_norm": 1.058761715888977,
|
| 722 |
+
"learning_rate": 1.8393514918970315e-05,
|
| 723 |
+
"loss": 0.4265,
|
| 724 |
+
"step": 510
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.3849259757738897,
|
| 728 |
+
"grad_norm": 1.1931772232055664,
|
| 729 |
+
"learning_rate": 1.8176630885862927e-05,
|
| 730 |
+
"loss": 0.452,
|
| 731 |
+
"step": 515
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.398384925975774,
|
| 735 |
+
"grad_norm": 1.1567202806472778,
|
| 736 |
+
"learning_rate": 1.7959049284305056e-05,
|
| 737 |
+
"loss": 0.4471,
|
| 738 |
+
"step": 520
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.4118438761776582,
|
| 742 |
+
"grad_norm": 1.1528855562210083,
|
| 743 |
+
"learning_rate": 1.7740817893867613e-05,
|
| 744 |
+
"loss": 0.4357,
|
| 745 |
+
"step": 525
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.4253028263795424,
|
| 749 |
+
"grad_norm": 1.1575230360031128,
|
| 750 |
+
"learning_rate": 1.752198463681111e-05,
|
| 751 |
+
"loss": 0.4249,
|
| 752 |
+
"step": 530
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.4387617765814267,
|
| 756 |
+
"grad_norm": 1.087706208229065,
|
| 757 |
+
"learning_rate": 1.7302597567562237e-05,
|
| 758 |
+
"loss": 0.4249,
|
| 759 |
+
"step": 535
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.452220726783311,
|
| 763 |
+
"grad_norm": 1.1799367666244507,
|
| 764 |
+
"learning_rate": 1.7082704862161422e-05,
|
| 765 |
+
"loss": 0.4086,
|
| 766 |
+
"step": 540
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.4656796769851952,
|
| 770 |
+
"grad_norm": 1.1974951028823853,
|
| 771 |
+
"learning_rate": 1.6862354807683643e-05,
|
| 772 |
+
"loss": 0.4195,
|
| 773 |
+
"step": 545
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.4791386271870794,
|
| 777 |
+
"grad_norm": 1.0563615560531616,
|
| 778 |
+
"learning_rate": 1.66415957916349e-05,
|
| 779 |
+
"loss": 0.4491,
|
| 780 |
+
"step": 550
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.4925975773889637,
|
| 784 |
+
"grad_norm": 1.0967886447906494,
|
| 785 |
+
"learning_rate": 1.642047629132663e-05,
|
| 786 |
+
"loss": 0.4067,
|
| 787 |
+
"step": 555
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.506056527590848,
|
| 791 |
+
"grad_norm": 1.0966241359710693,
|
| 792 |
+
"learning_rate": 1.619904486323037e-05,
|
| 793 |
+
"loss": 0.4362,
|
| 794 |
+
"step": 560
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.5195154777927322,
|
| 798 |
+
"grad_norm": 1.0399625301361084,
|
| 799 |
+
"learning_rate": 1.597735013231507e-05,
|
| 800 |
+
"loss": 0.4074,
|
| 801 |
+
"step": 565
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.5329744279946165,
|
| 805 |
+
"grad_norm": 0.995629072189331,
|
| 806 |
+
"learning_rate": 1.5755440781369345e-05,
|
| 807 |
+
"loss": 0.4502,
|
| 808 |
+
"step": 570
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.5464333781965007,
|
| 812 |
+
"grad_norm": 1.1102111339569092,
|
| 813 |
+
"learning_rate": 1.5533365540311038e-05,
|
| 814 |
+
"loss": 0.3681,
|
| 815 |
+
"step": 575
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.559892328398385,
|
| 819 |
+
"grad_norm": 1.2387511730194092,
|
| 820 |
+
"learning_rate": 1.531117317548643e-05,
|
| 821 |
+
"loss": 0.3995,
|
| 822 |
+
"step": 580
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.5733512786002692,
|
| 826 |
+
"grad_norm": 1.2061550617218018,
|
| 827 |
+
"learning_rate": 1.5088912478961458e-05,
|
| 828 |
+
"loss": 0.3832,
|
| 829 |
+
"step": 585
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.5868102288021535,
|
| 833 |
+
"grad_norm": 1.124131202697754,
|
| 834 |
+
"learning_rate": 1.4866632257807278e-05,
|
| 835 |
+
"loss": 0.3715,
|
| 836 |
+
"step": 590
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.6002691790040378,
|
| 840 |
+
"grad_norm": 1.1609008312225342,
|
| 841 |
+
"learning_rate": 1.4644381323382539e-05,
|
| 842 |
+
"loss": 0.4226,
|
| 843 |
+
"step": 595
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.613728129205922,
|
| 847 |
+
"grad_norm": 1.1971077919006348,
|
| 848 |
+
"learning_rate": 1.4422208480614684e-05,
|
| 849 |
+
"loss": 0.3835,
|
| 850 |
+
"step": 600
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.6271870794078063,
|
| 854 |
+
"grad_norm": 1.0978761911392212,
|
| 855 |
+
"learning_rate": 1.4200162517282736e-05,
|
| 856 |
+
"loss": 0.3972,
|
| 857 |
+
"step": 605
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.6406460296096905,
|
| 861 |
+
"grad_norm": 1.0205671787261963,
|
| 862 |
+
"learning_rate": 1.3978292193303768e-05,
|
| 863 |
+
"loss": 0.3835,
|
| 864 |
+
"step": 610
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.6541049798115748,
|
| 868 |
+
"grad_norm": 1.137688398361206,
|
| 869 |
+
"learning_rate": 1.3756646230025555e-05,
|
| 870 |
+
"loss": 0.4288,
|
| 871 |
+
"step": 615
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.667563930013459,
|
| 875 |
+
"grad_norm": 1.1691123247146606,
|
| 876 |
+
"learning_rate": 1.3535273299527622e-05,
|
| 877 |
+
"loss": 0.3641,
|
| 878 |
+
"step": 620
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.6810228802153433,
|
| 882 |
+
"grad_norm": 1.6706739664077759,
|
| 883 |
+
"learning_rate": 1.3314222013933218e-05,
|
| 884 |
+
"loss": 0.4099,
|
| 885 |
+
"step": 625
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.6944818304172276,
|
| 889 |
+
"grad_norm": 1.0793585777282715,
|
| 890 |
+
"learning_rate": 1.3093540914734351e-05,
|
| 891 |
+
"loss": 0.4181,
|
| 892 |
+
"step": 630
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.7079407806191118,
|
| 896 |
+
"grad_norm": 1.3619318008422852,
|
| 897 |
+
"learning_rate": 1.2873278462132401e-05,
|
| 898 |
+
"loss": 0.3841,
|
| 899 |
+
"step": 635
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.721399730820996,
|
| 903 |
+
"grad_norm": 1.095859169960022,
|
| 904 |
+
"learning_rate": 1.2653483024396535e-05,
|
| 905 |
+
"loss": 0.427,
|
| 906 |
+
"step": 640
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.7348586810228803,
|
| 910 |
+
"grad_norm": 1.0056387186050415,
|
| 911 |
+
"learning_rate": 1.2434202867242372e-05,
|
| 912 |
+
"loss": 0.3585,
|
| 913 |
+
"step": 645
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.7483176312247646,
|
| 917 |
+
"grad_norm": 1.1625523567199707,
|
| 918 |
+
"learning_rate": 1.221548614323308e-05,
|
| 919 |
+
"loss": 0.3616,
|
| 920 |
+
"step": 650
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.7617765814266488,
|
| 924 |
+
"grad_norm": 1.359569787979126,
|
| 925 |
+
"learning_rate": 1.1997380881205431e-05,
|
| 926 |
+
"loss": 0.333,
|
| 927 |
+
"step": 655
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.775235531628533,
|
| 931 |
+
"grad_norm": 1.1408214569091797,
|
| 932 |
+
"learning_rate": 1.1779934975722919e-05,
|
| 933 |
+
"loss": 0.3538,
|
| 934 |
+
"step": 660
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.7886944818304173,
|
| 938 |
+
"grad_norm": 1.2193180322647095,
|
| 939 |
+
"learning_rate": 1.1563196176558436e-05,
|
| 940 |
+
"loss": 0.3879,
|
| 941 |
+
"step": 665
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.8021534320323016,
|
| 945 |
+
"grad_norm": 1.2086423635482788,
|
| 946 |
+
"learning_rate": 1.13472120782087e-05,
|
| 947 |
+
"loss": 0.3682,
|
| 948 |
+
"step": 670
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.8156123822341859,
|
| 952 |
+
"grad_norm": 0.997719943523407,
|
| 953 |
+
"learning_rate": 1.1132030109442823e-05,
|
| 954 |
+
"loss": 0.3839,
|
| 955 |
+
"step": 675
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.8290713324360701,
|
| 959 |
+
"grad_norm": 1.1687936782836914,
|
| 960 |
+
"learning_rate": 1.0917697522887217e-05,
|
| 961 |
+
"loss": 0.3768,
|
| 962 |
+
"step": 680
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.8425302826379544,
|
| 966 |
+
"grad_norm": 1.1776505708694458,
|
| 967 |
+
"learning_rate": 1.0704261384649242e-05,
|
| 968 |
+
"loss": 0.346,
|
| 969 |
+
"step": 685
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.8559892328398386,
|
| 973 |
+
"grad_norm": 1.044433355331421,
|
| 974 |
+
"learning_rate": 1.0491768563981747e-05,
|
| 975 |
+
"loss": 0.3792,
|
| 976 |
+
"step": 690
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.8694481830417229,
|
| 980 |
+
"grad_norm": 0.9955368638038635,
|
| 981 |
+
"learning_rate": 1.0280265722990908e-05,
|
| 982 |
+
"loss": 0.3246,
|
| 983 |
+
"step": 695
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.8829071332436071,
|
| 987 |
+
"grad_norm": 1.0968743562698364,
|
| 988 |
+
"learning_rate": 1.0069799306389485e-05,
|
| 989 |
+
"loss": 0.3595,
|
| 990 |
+
"step": 700
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.8963660834454914,
|
| 994 |
+
"grad_norm": 1.1665064096450806,
|
| 995 |
+
"learning_rate": 9.860415531297881e-06,
|
| 996 |
+
"loss": 0.3774,
|
| 997 |
+
"step": 705
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.9098250336473757,
|
| 1001 |
+
"grad_norm": 1.1354666948318481,
|
| 1002 |
+
"learning_rate": 9.652160377095124e-06,
|
| 1003 |
+
"loss": 0.3582,
|
| 1004 |
+
"step": 710
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.92328398384926,
|
| 1008 |
+
"grad_norm": 1.1390712261199951,
|
| 1009 |
+
"learning_rate": 9.445079575322101e-06,
|
| 1010 |
+
"loss": 0.3594,
|
| 1011 |
+
"step": 715
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.9367429340511442,
|
| 1015 |
+
"grad_norm": 1.1198657751083374,
|
| 1016 |
+
"learning_rate": 9.239218599639171e-06,
|
| 1017 |
+
"loss": 0.3213,
|
| 1018 |
+
"step": 720
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.9502018842530284,
|
| 1022 |
+
"grad_norm": 1.2458171844482422,
|
| 1023 |
+
"learning_rate": 9.03462265584046e-06,
|
| 1024 |
+
"loss": 0.3293,
|
| 1025 |
+
"step": 725
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.9636608344549125,
|
| 1029 |
+
"grad_norm": 1.0921350717544556,
|
| 1030 |
+
"learning_rate": 8.831336671926924e-06,
|
| 1031 |
+
"loss": 0.3505,
|
| 1032 |
+
"step": 730
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.9771197846567967,
|
| 1036 |
+
"grad_norm": 1.1428606510162354,
|
| 1037 |
+
"learning_rate": 8.629405288240461e-06,
|
| 1038 |
+
"loss": 0.3357,
|
| 1039 |
+
"step": 735
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.990578734858681,
|
| 1043 |
+
"grad_norm": 1.0794802904129028,
|
| 1044 |
+
"learning_rate": 8.428872847661139e-06,
|
| 1045 |
+
"loss": 0.3505,
|
| 1046 |
+
"step": 740
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.0026917900403767,
|
| 1050 |
+
"grad_norm": 1.1085920333862305,
|
| 1051 |
+
"learning_rate": 8.229783385869807e-06,
|
| 1052 |
+
"loss": 0.2998,
|
| 1053 |
+
"step": 745
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.016150740242261,
|
| 1057 |
+
"grad_norm": 1.1450228691101074,
|
| 1058 |
+
"learning_rate": 8.03218062167811e-06,
|
| 1059 |
+
"loss": 0.2814,
|
| 1060 |
+
"step": 750
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.029609690444145,
|
| 1064 |
+
"grad_norm": 1.1784065961837769,
|
| 1065 |
+
"learning_rate": 7.836107947428115e-06,
|
| 1066 |
+
"loss": 0.2864,
|
| 1067 |
+
"step": 755
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.0430686406460294,
|
| 1071 |
+
"grad_norm": 1.024545431137085,
|
| 1072 |
+
"learning_rate": 7.641608419463621e-06,
|
| 1073 |
+
"loss": 0.2633,
|
| 1074 |
+
"step": 760
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.0565275908479137,
|
| 1078 |
+
"grad_norm": 1.183801531791687,
|
| 1079 |
+
"learning_rate": 7.448724748675273e-06,
|
| 1080 |
+
"loss": 0.2903,
|
| 1081 |
+
"step": 765
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.069986541049798,
|
| 1085 |
+
"grad_norm": 1.1893815994262695,
|
| 1086 |
+
"learning_rate": 7.257499291121473e-06,
|
| 1087 |
+
"loss": 0.2706,
|
| 1088 |
+
"step": 770
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.083445491251682,
|
| 1092 |
+
"grad_norm": 1.226550579071045,
|
| 1093 |
+
"learning_rate": 7.0679740387272896e-06,
|
| 1094 |
+
"loss": 0.2937,
|
| 1095 |
+
"step": 775
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.0969044414535665,
|
| 1099 |
+
"grad_norm": 1.2340624332427979,
|
| 1100 |
+
"learning_rate": 6.880190610063272e-06,
|
| 1101 |
+
"loss": 0.2722,
|
| 1102 |
+
"step": 780
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.1103633916554507,
|
| 1106 |
+
"grad_norm": 1.2975499629974365,
|
| 1107 |
+
"learning_rate": 6.694190241206277e-06,
|
| 1108 |
+
"loss": 0.258,
|
| 1109 |
+
"step": 785
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.123822341857335,
|
| 1113 |
+
"grad_norm": 1.0203527212142944,
|
| 1114 |
+
"learning_rate": 6.510013776684281e-06,
|
| 1115 |
+
"loss": 0.2509,
|
| 1116 |
+
"step": 790
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.1372812920592192,
|
| 1120 |
+
"grad_norm": 1.3438855409622192,
|
| 1121 |
+
"learning_rate": 6.327701660507191e-06,
|
| 1122 |
+
"loss": 0.2859,
|
| 1123 |
+
"step": 795
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.1507402422611035,
|
| 1127 |
+
"grad_norm": 1.202867865562439,
|
| 1128 |
+
"learning_rate": 6.147293927285537e-06,
|
| 1129 |
+
"loss": 0.264,
|
| 1130 |
+
"step": 800
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.1641991924629878,
|
| 1134 |
+
"grad_norm": 1.0623880624771118,
|
| 1135 |
+
"learning_rate": 5.968830193439195e-06,
|
| 1136 |
+
"loss": 0.3156,
|
| 1137 |
+
"step": 805
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.177658142664872,
|
| 1141 |
+
"grad_norm": 1.0969281196594238,
|
| 1142 |
+
"learning_rate": 5.7923496484978115e-06,
|
| 1143 |
+
"loss": 0.2448,
|
| 1144 |
+
"step": 810
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.1911170928667563,
|
| 1148 |
+
"grad_norm": 1.1712204217910767,
|
| 1149 |
+
"learning_rate": 5.617891046495046e-06,
|
| 1150 |
+
"loss": 0.2714,
|
| 1151 |
+
"step": 815
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.2045760430686405,
|
| 1155 |
+
"grad_norm": 1.1462839841842651,
|
| 1156 |
+
"learning_rate": 5.44549269745842e-06,
|
| 1157 |
+
"loss": 0.2491,
|
| 1158 |
+
"step": 820
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.218034993270525,
|
| 1162 |
+
"grad_norm": 1.1571354866027832,
|
| 1163 |
+
"learning_rate": 5.275192458996682e-06,
|
| 1164 |
+
"loss": 0.2679,
|
| 1165 |
+
"step": 825
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.231493943472409,
|
| 1169 |
+
"grad_norm": 1.1075127124786377,
|
| 1170 |
+
"learning_rate": 5.1070277279864926e-06,
|
| 1171 |
+
"loss": 0.2781,
|
| 1172 |
+
"step": 830
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.2449528936742933,
|
| 1176 |
+
"grad_norm": 1.0444560050964355,
|
| 1177 |
+
"learning_rate": 4.941035432360333e-06,
|
| 1178 |
+
"loss": 0.2483,
|
| 1179 |
+
"step": 835
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.2584118438761775,
|
| 1183 |
+
"grad_norm": 1.3500826358795166,
|
| 1184 |
+
"learning_rate": 4.777252022997327e-06,
|
| 1185 |
+
"loss": 0.2628,
|
| 1186 |
+
"step": 840
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.271870794078062,
|
| 1190 |
+
"grad_norm": 1.1868863105773926,
|
| 1191 |
+
"learning_rate": 4.615713465718867e-06,
|
| 1192 |
+
"loss": 0.2592,
|
| 1193 |
+
"step": 845
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.285329744279946,
|
| 1197 |
+
"grad_norm": 1.0396500825881958,
|
| 1198 |
+
"learning_rate": 4.456455233390751e-06,
|
| 1199 |
+
"loss": 0.2534,
|
| 1200 |
+
"step": 850
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.2987886944818303,
|
| 1204 |
+
"grad_norm": 1.1607229709625244,
|
| 1205 |
+
"learning_rate": 4.299512298133546e-06,
|
| 1206 |
+
"loss": 0.2782,
|
| 1207 |
+
"step": 855
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.3122476446837146,
|
| 1211 |
+
"grad_norm": 1.3405728340148926,
|
| 1212 |
+
"learning_rate": 4.144919123642947e-06,
|
| 1213 |
+
"loss": 0.2784,
|
| 1214 |
+
"step": 860
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.325706594885599,
|
| 1218 |
+
"grad_norm": 1.042685627937317,
|
| 1219 |
+
"learning_rate": 3.992709657621739e-06,
|
| 1220 |
+
"loss": 0.248,
|
| 1221 |
+
"step": 865
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.339165545087483,
|
| 1225 |
+
"grad_norm": 1.0927951335906982,
|
| 1226 |
+
"learning_rate": 3.842917324325107e-06,
|
| 1227 |
+
"loss": 0.2642,
|
| 1228 |
+
"step": 870
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.3526244952893673,
|
| 1232 |
+
"grad_norm": 1.212036371231079,
|
| 1233 |
+
"learning_rate": 3.6955750172208763e-06,
|
| 1234 |
+
"loss": 0.2477,
|
| 1235 |
+
"step": 875
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.3660834454912516,
|
| 1239 |
+
"grad_norm": 1.1431571245193481,
|
| 1240 |
+
"learning_rate": 3.550715091766307e-06,
|
| 1241 |
+
"loss": 0.2751,
|
| 1242 |
+
"step": 880
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.379542395693136,
|
| 1246 |
+
"grad_norm": 1.0427595376968384,
|
| 1247 |
+
"learning_rate": 3.4083693583030306e-06,
|
| 1248 |
+
"loss": 0.2686,
|
| 1249 |
+
"step": 885
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.39300134589502,
|
| 1253 |
+
"grad_norm": 1.1862140893936157,
|
| 1254 |
+
"learning_rate": 3.268569075071722e-06,
|
| 1255 |
+
"loss": 0.2837,
|
| 1256 |
+
"step": 890
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.4064602960969044,
|
| 1260 |
+
"grad_norm": 1.0857174396514893,
|
| 1261 |
+
"learning_rate": 3.131344941347937e-06,
|
| 1262 |
+
"loss": 0.2632,
|
| 1263 |
+
"step": 895
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.4199192462987886,
|
| 1267 |
+
"grad_norm": 1.2340177297592163,
|
| 1268 |
+
"learning_rate": 2.996727090700794e-06,
|
| 1269 |
+
"loss": 0.2635,
|
| 1270 |
+
"step": 900
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.433378196500673,
|
| 1274 |
+
"grad_norm": 1.1067768335342407,
|
| 1275 |
+
"learning_rate": 2.86474508437579e-06,
|
| 1276 |
+
"loss": 0.2617,
|
| 1277 |
+
"step": 905
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.446837146702557,
|
| 1281 |
+
"grad_norm": 1.1215053796768188,
|
| 1282 |
+
"learning_rate": 2.7354279048033524e-06,
|
| 1283 |
+
"loss": 0.2606,
|
| 1284 |
+
"step": 910
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.4602960969044414,
|
| 1288 |
+
"grad_norm": 1.1256719827651978,
|
| 1289 |
+
"learning_rate": 2.6088039492344707e-06,
|
| 1290 |
+
"loss": 0.2487,
|
| 1291 |
+
"step": 915
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.4737550471063257,
|
| 1295 |
+
"grad_norm": 1.2089550495147705,
|
| 1296 |
+
"learning_rate": 2.4849010235048513e-06,
|
| 1297 |
+
"loss": 0.2404,
|
| 1298 |
+
"step": 920
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.48721399730821,
|
| 1302 |
+
"grad_norm": 1.1536446809768677,
|
| 1303 |
+
"learning_rate": 2.3637463359288914e-06,
|
| 1304 |
+
"loss": 0.2633,
|
| 1305 |
+
"step": 925
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.500672947510094,
|
| 1309 |
+
"grad_norm": 1.0565274953842163,
|
| 1310 |
+
"learning_rate": 2.2453664913249505e-06,
|
| 1311 |
+
"loss": 0.2311,
|
| 1312 |
+
"step": 930
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.5141318977119784,
|
| 1316 |
+
"grad_norm": 1.1993619203567505,
|
| 1317 |
+
"learning_rate": 2.1297874851730575e-06,
|
| 1318 |
+
"loss": 0.274,
|
| 1319 |
+
"step": 935
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.5275908479138627,
|
| 1323 |
+
"grad_norm": 1.0851179361343384,
|
| 1324 |
+
"learning_rate": 2.017034697906489e-06,
|
| 1325 |
+
"loss": 0.2326,
|
| 1326 |
+
"step": 940
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.541049798115747,
|
| 1330 |
+
"grad_norm": 1.1467525959014893,
|
| 1331 |
+
"learning_rate": 1.9071328893383667e-06,
|
| 1332 |
+
"loss": 0.242,
|
| 1333 |
+
"step": 945
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.554508748317631,
|
| 1337 |
+
"grad_norm": 1.3856874704360962,
|
| 1338 |
+
"learning_rate": 1.8001061932245654e-06,
|
| 1339 |
+
"loss": 0.2445,
|
| 1340 |
+
"step": 950
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.5679676985195155,
|
| 1344 |
+
"grad_norm": 1.1563737392425537,
|
| 1345 |
+
"learning_rate": 1.6959781119640894e-06,
|
| 1346 |
+
"loss": 0.2311,
|
| 1347 |
+
"step": 955
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.5814266487213997,
|
| 1351 |
+
"grad_norm": 1.0257714986801147,
|
| 1352 |
+
"learning_rate": 1.594771511438095e-06,
|
| 1353 |
+
"loss": 0.2504,
|
| 1354 |
+
"step": 960
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.594885598923284,
|
| 1358 |
+
"grad_norm": 1.142681360244751,
|
| 1359 |
+
"learning_rate": 1.4965086159886694e-06,
|
| 1360 |
+
"loss": 0.2349,
|
| 1361 |
+
"step": 965
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.608344549125168,
|
| 1365 |
+
"grad_norm": 1.1093924045562744,
|
| 1366 |
+
"learning_rate": 1.401211003538519e-06,
|
| 1367 |
+
"loss": 0.2381,
|
| 1368 |
+
"step": 970
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.6218034993270525,
|
| 1372 |
+
"grad_norm": 1.2210307121276855,
|
| 1373 |
+
"learning_rate": 1.308899600852585e-06,
|
| 1374 |
+
"loss": 0.2722,
|
| 1375 |
+
"step": 975
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.6352624495289367,
|
| 1379 |
+
"grad_norm": 1.1265455484390259,
|
| 1380 |
+
"learning_rate": 1.2195946789426531e-06,
|
| 1381 |
+
"loss": 0.2578,
|
| 1382 |
+
"step": 980
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.648721399730821,
|
| 1386 |
+
"grad_norm": 1.0628033876419067,
|
| 1387 |
+
"learning_rate": 1.13331584861597e-06,
|
| 1388 |
+
"loss": 0.2182,
|
| 1389 |
+
"step": 985
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.6621803499327052,
|
| 1393 |
+
"grad_norm": 1.1813066005706787,
|
| 1394 |
+
"learning_rate": 1.0500820561688374e-06,
|
| 1395 |
+
"loss": 0.2619,
|
| 1396 |
+
"step": 990
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.6756393001345895,
|
| 1400 |
+
"grad_norm": 1.1125173568725586,
|
| 1401 |
+
"learning_rate": 9.699115792260953e-07,
|
| 1402 |
+
"loss": 0.2933,
|
| 1403 |
+
"step": 995
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 2.6890982503364738,
|
| 1407 |
+
"grad_norm": 1.0151184797286987,
|
| 1408 |
+
"learning_rate": 8.928220227275086e-07,
|
| 1409 |
+
"loss": 0.2342,
|
| 1410 |
+
"step": 1000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.702557200538358,
|
| 1414 |
+
"grad_norm": 0.9034965634346008,
|
| 1415 |
+
"learning_rate": 8.18830315061801e-07,
|
| 1416 |
+
"loss": 0.2455,
|
| 1417 |
+
"step": 1005
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.7160161507402423,
|
| 1421 |
+
"grad_norm": 1.1290091276168823,
|
| 1422 |
+
"learning_rate": 7.479527043492984e-07,
|
| 1423 |
+
"loss": 0.237,
|
| 1424 |
+
"step": 1010
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.7294751009421265,
|
| 1428 |
+
"grad_norm": 1.0831630229949951,
|
| 1429 |
+
"learning_rate": 6.802047548739409e-07,
|
| 1430 |
+
"loss": 0.2242,
|
| 1431 |
+
"step": 1015
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.742934051144011,
|
| 1435 |
+
"grad_norm": 1.130257248878479,
|
| 1436 |
+
"learning_rate": 6.156013436654617e-07,
|
| 1437 |
+
"loss": 0.2319,
|
| 1438 |
+
"step": 1020
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.756393001345895,
|
| 1442 |
+
"grad_norm": 1.158980369567871,
|
| 1443 |
+
"learning_rate": 5.541566572324786e-07,
|
| 1444 |
+
"loss": 0.263,
|
| 1445 |
+
"step": 1025
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.7698519515477793,
|
| 1449 |
+
"grad_norm": 1.0408811569213867,
|
| 1450 |
+
"learning_rate": 4.958841884472409e-07,
|
| 1451 |
+
"loss": 0.2423,
|
| 1452 |
+
"step": 1030
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.7833109017496636,
|
| 1456 |
+
"grad_norm": 1.1301546096801758,
|
| 1457 |
+
"learning_rate": 4.407967335826585e-07,
|
| 1458 |
+
"loss": 0.2513,
|
| 1459 |
+
"step": 1035
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.796769851951548,
|
| 1463 |
+
"grad_norm": 1.0434672832489014,
|
| 1464 |
+
"learning_rate": 3.889063895023287e-07,
|
| 1465 |
+
"loss": 0.2385,
|
| 1466 |
+
"step": 1040
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.810228802153432,
|
| 1470 |
+
"grad_norm": 1.2883914709091187,
|
| 1471 |
+
"learning_rate": 3.4022455100414184e-07,
|
| 1472 |
+
"loss": 0.2629,
|
| 1473 |
+
"step": 1045
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.8236877523553163,
|
| 1477 |
+
"grad_norm": 1.1323546171188354,
|
| 1478 |
+
"learning_rate": 2.947619083180525e-07,
|
| 1479 |
+
"loss": 0.251,
|
| 1480 |
+
"step": 1050
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.8371467025572006,
|
| 1484 |
+
"grad_norm": 1.0120676755905151,
|
| 1485 |
+
"learning_rate": 2.5252844475856906e-07,
|
| 1486 |
+
"loss": 0.2367,
|
| 1487 |
+
"step": 1055
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.850605652759085,
|
| 1491 |
+
"grad_norm": 1.0989749431610107,
|
| 1492 |
+
"learning_rate": 2.1353343453248408e-07,
|
| 1493 |
+
"loss": 0.2462,
|
| 1494 |
+
"step": 1060
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.864064602960969,
|
| 1498 |
+
"grad_norm": 0.9685646295547485,
|
| 1499 |
+
"learning_rate": 1.777854407023083e-07,
|
| 1500 |
+
"loss": 0.2408,
|
| 1501 |
+
"step": 1065
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.8775235531628534,
|
| 1505 |
+
"grad_norm": 1.096001148223877,
|
| 1506 |
+
"learning_rate": 1.4529231330588644e-07,
|
| 1507 |
+
"loss": 0.2486,
|
| 1508 |
+
"step": 1070
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.8909825033647376,
|
| 1512 |
+
"grad_norm": 1.1372036933898926,
|
| 1513 |
+
"learning_rate": 1.160611876325679e-07,
|
| 1514 |
+
"loss": 0.2621,
|
| 1515 |
+
"step": 1075
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.904441453566622,
|
| 1519 |
+
"grad_norm": 1.0930324792861938,
|
| 1520 |
+
"learning_rate": 9.009848265634669e-08,
|
| 1521 |
+
"loss": 0.2618,
|
| 1522 |
+
"step": 1080
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.917900403768506,
|
| 1526 |
+
"grad_norm": 0.9669865369796753,
|
| 1527 |
+
"learning_rate": 6.740989962628896e-08,
|
| 1528 |
+
"loss": 0.227,
|
| 1529 |
+
"step": 1085
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.9313593539703904,
|
| 1533 |
+
"grad_norm": 1.0694055557250977,
|
| 1534 |
+
"learning_rate": 4.800042081458456e-08,
|
| 1535 |
+
"loss": 0.2678,
|
| 1536 |
+
"step": 1090
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.9448183041722746,
|
| 1540 |
+
"grad_norm": 1.0851000547409058,
|
| 1541 |
+
"learning_rate": 3.187430842245709e-08,
|
| 1542 |
+
"loss": 0.2118,
|
| 1543 |
+
"step": 1095
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.958277254374159,
|
| 1547 |
+
"grad_norm": 1.1466342210769653,
|
| 1548 |
+
"learning_rate": 1.9035103644222675e-08,
|
| 1549 |
+
"loss": 0.2271,
|
| 1550 |
+
"step": 1100
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.971736204576043,
|
| 1554 |
+
"grad_norm": 1.0995334386825562,
|
| 1555 |
+
"learning_rate": 9.485625889660288e-09,
|
| 1556 |
+
"loss": 0.2574,
|
| 1557 |
+
"step": 1105
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.9851951547779274,
|
| 1561 |
+
"grad_norm": 1.0613255500793457,
|
| 1562 |
+
"learning_rate": 3.2279721648870162e-09,
|
| 1563 |
+
"loss": 0.2179,
|
| 1564 |
+
"step": 1110
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.9986541049798117,
|
| 1568 |
+
"grad_norm": 1.0960551500320435,
|
| 1569 |
+
"learning_rate": 2.6351661186974074e-10,
|
| 1570 |
+
"loss": 0.2189,
|
| 1571 |
+
"step": 1115
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 3.0,
|
| 1575 |
+
"step": 1116,
|
| 1576 |
+
"total_flos": 1.5875073592094884e+18,
|
| 1577 |
+
"train_loss": 0.5184633527551928,
|
| 1578 |
+
"train_runtime": 742.8841,
|
| 1579 |
+
"train_samples_per_second": 47.963,
|
| 1580 |
+
"train_steps_per_second": 1.502
|
| 1581 |
+
}
|
| 1582 |
+
],
|
| 1583 |
+
"logging_steps": 5,
|
| 1584 |
+
"max_steps": 1116,
|
| 1585 |
+
"num_input_tokens_seen": 0,
|
| 1586 |
+
"num_train_epochs": 3,
|
| 1587 |
+
"save_steps": 2000,
|
| 1588 |
+
"stateful_callbacks": {
|
| 1589 |
+
"TrainerControl": {
|
| 1590 |
+
"args": {
|
| 1591 |
+
"should_epoch_stop": false,
|
| 1592 |
+
"should_evaluate": false,
|
| 1593 |
+
"should_log": false,
|
| 1594 |
+
"should_save": false,
|
| 1595 |
+
"should_training_stop": false
|
| 1596 |
+
},
|
| 1597 |
+
"attributes": {}
|
| 1598 |
+
}
|
| 1599 |
+
},
|
| 1600 |
+
"total_flos": 1.5875073592094884e+18,
|
| 1601 |
+
"train_batch_size": 2,
|
| 1602 |
+
"trial_name": null,
|
| 1603 |
+
"trial_params": null
|
| 1604 |
+
}
|
49_128_e3_3e-5/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a96fa255e81544f1dc2ef41869395efbe9db155f4a3a73a7d1fe297e1f7a706a
|
| 3 |
+
size 8145
|
49_128_e3_3e-5/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|