Upload folder using huggingface_hub
Browse files- 4_128_e3_3e-5/README.md +63 -0
- 4_128_e3_3e-5/adapter_config.json +39 -0
- 4_128_e3_3e-5/adapter_model.safetensors +3 -0
- 4_128_e3_3e-5/all_results.json +9 -0
- 4_128_e3_3e-5/config.json +32 -0
- 4_128_e3_3e-5/merges.txt +0 -0
- 4_128_e3_3e-5/special_tokens_map.json +45 -0
- 4_128_e3_3e-5/tokenizer.json +0 -0
- 4_128_e3_3e-5/tokenizer_config.json +188 -0
- 4_128_e3_3e-5/train_results.json +9 -0
- 4_128_e3_3e-5/trainer_state.json +1604 -0
- 4_128_e3_3e-5/training_args.bin +3 -0
- 4_128_e3_3e-5/vocab.json +0 -0
4_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: 4_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 |
+
# 4_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
|
4_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 |
+
"up_proj",
|
| 28 |
+
"q_proj",
|
| 29 |
+
"v_proj",
|
| 30 |
+
"k_proj",
|
| 31 |
+
"o_proj",
|
| 32 |
+
"down_proj",
|
| 33 |
+
"gate_proj"
|
| 34 |
+
],
|
| 35 |
+
"task_type": "CAUSAL_LM",
|
| 36 |
+
"trainable_token_indices": null,
|
| 37 |
+
"use_dora": false,
|
| 38 |
+
"use_rslora": false
|
| 39 |
+
}
|
4_128_e3_3e-5/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4db078d940fc3e313688dfd11fae025731ae19a2ef462d7ba1e35a87c5add446
|
| 3 |
+
size 791751704
|
4_128_e3_3e-5/all_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.647805420300927e+18,
|
| 4 |
+
"train_loss": 0.5340931572159877,
|
| 5 |
+
"train_runtime": 781.1418,
|
| 6 |
+
"train_samples": 11931,
|
| 7 |
+
"train_samples_per_second": 45.821,
|
| 8 |
+
"train_steps_per_second": 1.433
|
| 9 |
+
}
|
4_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 |
+
}
|
4_128_e3_3e-5/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
4_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 |
+
}
|
4_128_e3_3e-5/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
4_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 |
+
}
|
4_128_e3_3e-5/train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.647805420300927e+18,
|
| 4 |
+
"train_loss": 0.5340931572159877,
|
| 5 |
+
"train_runtime": 781.1418,
|
| 6 |
+
"train_samples": 11931,
|
| 7 |
+
"train_samples_per_second": 45.821,
|
| 8 |
+
"train_steps_per_second": 1.433
|
| 9 |
+
}
|
4_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": 1119,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.013404825737265416,
|
| 14 |
+
"grad_norm": 1.2472357749938965,
|
| 15 |
+
"learning_rate": 2.1428571428571427e-06,
|
| 16 |
+
"loss": 1.3007,
|
| 17 |
+
"step": 5
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02680965147453083,
|
| 21 |
+
"grad_norm": 1.018729567527771,
|
| 22 |
+
"learning_rate": 4.821428571428572e-06,
|
| 23 |
+
"loss": 1.2872,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.040214477211796246,
|
| 28 |
+
"grad_norm": 0.6834114193916321,
|
| 29 |
+
"learning_rate": 7.5e-06,
|
| 30 |
+
"loss": 1.2708,
|
| 31 |
+
"step": 15
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.05361930294906166,
|
| 35 |
+
"grad_norm": 0.924774706363678,
|
| 36 |
+
"learning_rate": 1.0178571428571429e-05,
|
| 37 |
+
"loss": 1.1525,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.06702412868632708,
|
| 42 |
+
"grad_norm": 0.5801697373390198,
|
| 43 |
+
"learning_rate": 1.2857142857142857e-05,
|
| 44 |
+
"loss": 1.2542,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.08042895442359249,
|
| 49 |
+
"grad_norm": 0.5304354429244995,
|
| 50 |
+
"learning_rate": 1.553571428571429e-05,
|
| 51 |
+
"loss": 1.1908,
|
| 52 |
+
"step": 30
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0938337801608579,
|
| 56 |
+
"grad_norm": 0.48565009236335754,
|
| 57 |
+
"learning_rate": 1.8214285714285712e-05,
|
| 58 |
+
"loss": 1.1505,
|
| 59 |
+
"step": 35
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.10723860589812333,
|
| 63 |
+
"grad_norm": 0.46631884574890137,
|
| 64 |
+
"learning_rate": 2.089285714285714e-05,
|
| 65 |
+
"loss": 1.2448,
|
| 66 |
+
"step": 40
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.12064343163538874,
|
| 70 |
+
"grad_norm": 0.600545346736908,
|
| 71 |
+
"learning_rate": 2.357142857142857e-05,
|
| 72 |
+
"loss": 1.1581,
|
| 73 |
+
"step": 45
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.13404825737265416,
|
| 77 |
+
"grad_norm": 0.527691662311554,
|
| 78 |
+
"learning_rate": 2.625e-05,
|
| 79 |
+
"loss": 1.1731,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.14745308310991956,
|
| 84 |
+
"grad_norm": 0.48526814579963684,
|
| 85 |
+
"learning_rate": 2.892857142857143e-05,
|
| 86 |
+
"loss": 1.1426,
|
| 87 |
+
"step": 55
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.16085790884718498,
|
| 91 |
+
"grad_norm": 0.410771906375885,
|
| 92 |
+
"learning_rate": 2.999941043167295e-05,
|
| 93 |
+
"loss": 1.1407,
|
| 94 |
+
"step": 60
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.1742627345844504,
|
| 98 |
+
"grad_norm": 0.5673904418945312,
|
| 99 |
+
"learning_rate": 2.999580768195271e-05,
|
| 100 |
+
"loss": 1.0899,
|
| 101 |
+
"step": 65
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.1876675603217158,
|
| 105 |
+
"grad_norm": 0.5336377024650574,
|
| 106 |
+
"learning_rate": 2.998893050620046e-05,
|
| 107 |
+
"loss": 1.0582,
|
| 108 |
+
"step": 70
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.20107238605898123,
|
| 112 |
+
"grad_norm": 0.53374844789505,
|
| 113 |
+
"learning_rate": 2.9978780406089445e-05,
|
| 114 |
+
"loss": 1.0804,
|
| 115 |
+
"step": 75
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.21447721179624665,
|
| 119 |
+
"grad_norm": 0.4427356421947479,
|
| 120 |
+
"learning_rate": 2.996535959795591e-05,
|
| 121 |
+
"loss": 1.0686,
|
| 122 |
+
"step": 80
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.22788203753351208,
|
| 126 |
+
"grad_norm": 0.48419925570487976,
|
| 127 |
+
"learning_rate": 2.994867101231513e-05,
|
| 128 |
+
"loss": 1.0867,
|
| 129 |
+
"step": 85
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.24128686327077747,
|
| 133 |
+
"grad_norm": 0.5140860676765442,
|
| 134 |
+
"learning_rate": 2.9928718293221532e-05,
|
| 135 |
+
"loss": 1.0337,
|
| 136 |
+
"step": 90
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.2546916890080429,
|
| 140 |
+
"grad_norm": 0.5096598863601685,
|
| 141 |
+
"learning_rate": 2.9905505797472965e-05,
|
| 142 |
+
"loss": 1.0434,
|
| 143 |
+
"step": 95
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.2680965147453083,
|
| 147 |
+
"grad_norm": 0.5465592741966248,
|
| 148 |
+
"learning_rate": 2.9879038593659403e-05,
|
| 149 |
+
"loss": 1.0407,
|
| 150 |
+
"step": 100
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.28150134048257375,
|
| 154 |
+
"grad_norm": 0.6191142797470093,
|
| 155 |
+
"learning_rate": 2.984932246105616e-05,
|
| 156 |
+
"loss": 1.0505,
|
| 157 |
+
"step": 105
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.2949061662198391,
|
| 161 |
+
"grad_norm": 0.5154615640640259,
|
| 162 |
+
"learning_rate": 2.981636388836196e-05,
|
| 163 |
+
"loss": 0.9597,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.30831099195710454,
|
| 168 |
+
"grad_norm": 0.6329454183578491,
|
| 169 |
+
"learning_rate": 2.978017007228208e-05,
|
| 170 |
+
"loss": 0.9858,
|
| 171 |
+
"step": 115
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.32171581769436997,
|
| 175 |
+
"grad_norm": 0.575188934803009,
|
| 176 |
+
"learning_rate": 2.9740748915956932e-05,
|
| 177 |
+
"loss": 0.9933,
|
| 178 |
+
"step": 120
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3351206434316354,
|
| 182 |
+
"grad_norm": 0.502245306968689,
|
| 183 |
+
"learning_rate": 2.9698109027236335e-05,
|
| 184 |
+
"loss": 0.9603,
|
| 185 |
+
"step": 125
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.3485254691689008,
|
| 189 |
+
"grad_norm": 0.5507754683494568,
|
| 190 |
+
"learning_rate": 2.965225971679996e-05,
|
| 191 |
+
"loss": 0.981,
|
| 192 |
+
"step": 130
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.36193029490616624,
|
| 196 |
+
"grad_norm": 0.7022887468338013,
|
| 197 |
+
"learning_rate": 2.9603210996124257e-05,
|
| 198 |
+
"loss": 0.9451,
|
| 199 |
+
"step": 135
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.3753351206434316,
|
| 203 |
+
"grad_norm": 0.611986517906189,
|
| 204 |
+
"learning_rate": 2.9550973575296424e-05,
|
| 205 |
+
"loss": 0.9538,
|
| 206 |
+
"step": 140
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.38873994638069703,
|
| 210 |
+
"grad_norm": 0.8291884064674377,
|
| 211 |
+
"learning_rate": 2.949555886067578e-05,
|
| 212 |
+
"loss": 0.995,
|
| 213 |
+
"step": 145
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.40214477211796246,
|
| 217 |
+
"grad_norm": 0.6599870324134827,
|
| 218 |
+
"learning_rate": 2.9436978952403103e-05,
|
| 219 |
+
"loss": 0.8675,
|
| 220 |
+
"step": 150
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.4155495978552279,
|
| 224 |
+
"grad_norm": 0.6758149266242981,
|
| 225 |
+
"learning_rate": 2.937524664175851e-05,
|
| 226 |
+
"loss": 0.9174,
|
| 227 |
+
"step": 155
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.4289544235924933,
|
| 231 |
+
"grad_norm": 0.6708205938339233,
|
| 232 |
+
"learning_rate": 2.931037540836839e-05,
|
| 233 |
+
"loss": 0.8959,
|
| 234 |
+
"step": 160
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.44235924932975873,
|
| 238 |
+
"grad_norm": 0.7568298578262329,
|
| 239 |
+
"learning_rate": 2.9242379417262052e-05,
|
| 240 |
+
"loss": 0.9291,
|
| 241 |
+
"step": 165
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.45576407506702415,
|
| 245 |
+
"grad_norm": 0.7548916339874268,
|
| 246 |
+
"learning_rate": 2.9171273515778703e-05,
|
| 247 |
+
"loss": 0.946,
|
| 248 |
+
"step": 170
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.4691689008042895,
|
| 252 |
+
"grad_norm": 0.737224280834198,
|
| 253 |
+
"learning_rate": 2.909707323032545e-05,
|
| 254 |
+
"loss": 0.8876,
|
| 255 |
+
"step": 175
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.48257372654155495,
|
| 259 |
+
"grad_norm": 0.7204071879386902,
|
| 260 |
+
"learning_rate": 2.901979476298699e-05,
|
| 261 |
+
"loss": 0.8444,
|
| 262 |
+
"step": 180
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.4959785522788204,
|
| 266 |
+
"grad_norm": 0.794966995716095,
|
| 267 |
+
"learning_rate": 2.893945498798781e-05,
|
| 268 |
+
"loss": 0.8443,
|
| 269 |
+
"step": 185
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.5093833780160858,
|
| 273 |
+
"grad_norm": 0.8569888472557068,
|
| 274 |
+
"learning_rate": 2.885607144800759e-05,
|
| 275 |
+
"loss": 0.8748,
|
| 276 |
+
"step": 190
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.5227882037533512,
|
| 280 |
+
"grad_norm": 0.708555281162262,
|
| 281 |
+
"learning_rate": 2.876966235035064e-05,
|
| 282 |
+
"loss": 0.8526,
|
| 283 |
+
"step": 195
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.5361930294906166,
|
| 287 |
+
"grad_norm": 0.7718046307563782,
|
| 288 |
+
"learning_rate": 2.8680246562970253e-05,
|
| 289 |
+
"loss": 0.8597,
|
| 290 |
+
"step": 200
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.5495978552278821,
|
| 294 |
+
"grad_norm": 0.8034407496452332,
|
| 295 |
+
"learning_rate": 2.8587843610348735e-05,
|
| 296 |
+
"loss": 0.8829,
|
| 297 |
+
"step": 205
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.5630026809651475,
|
| 301 |
+
"grad_norm": 0.8720042705535889,
|
| 302 |
+
"learning_rate": 2.8492473669234143e-05,
|
| 303 |
+
"loss": 0.8171,
|
| 304 |
+
"step": 210
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.5764075067024129,
|
| 308 |
+
"grad_norm": 0.851708173751831,
|
| 309 |
+
"learning_rate": 2.8394157564234564e-05,
|
| 310 |
+
"loss": 0.8083,
|
| 311 |
+
"step": 215
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.5898123324396782,
|
| 315 |
+
"grad_norm": 0.7046942710876465,
|
| 316 |
+
"learning_rate": 2.8292916763270918e-05,
|
| 317 |
+
"loss": 0.8058,
|
| 318 |
+
"step": 220
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.6032171581769437,
|
| 322 |
+
"grad_norm": 0.7370072603225708,
|
| 323 |
+
"learning_rate": 2.818877337288934e-05,
|
| 324 |
+
"loss": 0.7986,
|
| 325 |
+
"step": 225
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.6166219839142091,
|
| 329 |
+
"grad_norm": 0.8055124878883362,
|
| 330 |
+
"learning_rate": 2.808175013343404e-05,
|
| 331 |
+
"loss": 0.8417,
|
| 332 |
+
"step": 230
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.6300268096514745,
|
| 336 |
+
"grad_norm": 0.9540407061576843,
|
| 337 |
+
"learning_rate": 2.7971870414081845e-05,
|
| 338 |
+
"loss": 0.7665,
|
| 339 |
+
"step": 235
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.6434316353887399,
|
| 343 |
+
"grad_norm": 0.869381308555603,
|
| 344 |
+
"learning_rate": 2.785915820773937e-05,
|
| 345 |
+
"loss": 0.8064,
|
| 346 |
+
"step": 240
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.6568364611260054,
|
| 350 |
+
"grad_norm": 0.8873690366744995,
|
| 351 |
+
"learning_rate": 2.774363812580405e-05,
|
| 352 |
+
"loss": 0.7943,
|
| 353 |
+
"step": 245
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.6702412868632708,
|
| 357 |
+
"grad_norm": 0.8190541863441467,
|
| 358 |
+
"learning_rate": 2.7625335392790056e-05,
|
| 359 |
+
"loss": 0.7807,
|
| 360 |
+
"step": 250
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.6836461126005362,
|
| 364 |
+
"grad_norm": 0.7668455839157104,
|
| 365 |
+
"learning_rate": 2.75042758408204e-05,
|
| 366 |
+
"loss": 0.7612,
|
| 367 |
+
"step": 255
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.6970509383378016,
|
| 371 |
+
"grad_norm": 0.925933837890625,
|
| 372 |
+
"learning_rate": 2.7380485903986317e-05,
|
| 373 |
+
"loss": 0.8064,
|
| 374 |
+
"step": 260
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.710455764075067,
|
| 378 |
+
"grad_norm": 0.8035215139389038,
|
| 379 |
+
"learning_rate": 2.7253992612575225e-05,
|
| 380 |
+
"loss": 0.7501,
|
| 381 |
+
"step": 265
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.7238605898123325,
|
| 385 |
+
"grad_norm": 0.8842463493347168,
|
| 386 |
+
"learning_rate": 2.712482358716848e-05,
|
| 387 |
+
"loss": 0.7334,
|
| 388 |
+
"step": 270
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.7372654155495979,
|
| 392 |
+
"grad_norm": 1.0536301136016846,
|
| 393 |
+
"learning_rate": 2.6993007032610297e-05,
|
| 394 |
+
"loss": 0.7069,
|
| 395 |
+
"step": 275
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.7506702412868632,
|
| 399 |
+
"grad_norm": 0.8780192732810974,
|
| 400 |
+
"learning_rate": 2.6858571731848997e-05,
|
| 401 |
+
"loss": 0.727,
|
| 402 |
+
"step": 280
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.7640750670241286,
|
| 406 |
+
"grad_norm": 0.8872283101081848,
|
| 407 |
+
"learning_rate": 2.672154703965212e-05,
|
| 408 |
+
"loss": 0.733,
|
| 409 |
+
"step": 285
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.7774798927613941,
|
| 413 |
+
"grad_norm": 0.8710986375808716,
|
| 414 |
+
"learning_rate": 2.6581962876196593e-05,
|
| 415 |
+
"loss": 0.6887,
|
| 416 |
+
"step": 290
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.7908847184986595,
|
| 420 |
+
"grad_norm": 0.8486837148666382,
|
| 421 |
+
"learning_rate": 2.643984972053551e-05,
|
| 422 |
+
"loss": 0.7331,
|
| 423 |
+
"step": 295
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.8042895442359249,
|
| 427 |
+
"grad_norm": 1.0940479040145874,
|
| 428 |
+
"learning_rate": 2.6295238603942832e-05,
|
| 429 |
+
"loss": 0.6835,
|
| 430 |
+
"step": 300
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.8176943699731903,
|
| 434 |
+
"grad_norm": 0.8243003487586975,
|
| 435 |
+
"learning_rate": 2.6148161103137515e-05,
|
| 436 |
+
"loss": 0.7032,
|
| 437 |
+
"step": 305
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.8310991957104558,
|
| 441 |
+
"grad_norm": 0.9313507676124573,
|
| 442 |
+
"learning_rate": 2.599864933338854e-05,
|
| 443 |
+
"loss": 0.686,
|
| 444 |
+
"step": 310
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.8445040214477212,
|
| 448 |
+
"grad_norm": 0.8665029406547546,
|
| 449 |
+
"learning_rate": 2.5846735941502356e-05,
|
| 450 |
+
"loss": 0.6883,
|
| 451 |
+
"step": 315
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.8579088471849866,
|
| 455 |
+
"grad_norm": 1.0570703744888306,
|
| 456 |
+
"learning_rate": 2.5692454098694256e-05,
|
| 457 |
+
"loss": 0.69,
|
| 458 |
+
"step": 320
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.871313672922252,
|
| 462 |
+
"grad_norm": 1.0085052251815796,
|
| 463 |
+
"learning_rate": 2.553583749334522e-05,
|
| 464 |
+
"loss": 0.7056,
|
| 465 |
+
"step": 325
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.8847184986595175,
|
| 469 |
+
"grad_norm": 0.9562423825263977,
|
| 470 |
+
"learning_rate": 2.537692032364587e-05,
|
| 471 |
+
"loss": 0.678,
|
| 472 |
+
"step": 330
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.8981233243967829,
|
| 476 |
+
"grad_norm": 0.9648351669311523,
|
| 477 |
+
"learning_rate": 2.521573729012907e-05,
|
| 478 |
+
"loss": 0.6541,
|
| 479 |
+
"step": 335
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.9115281501340483,
|
| 483 |
+
"grad_norm": 1.0201152563095093,
|
| 484 |
+
"learning_rate": 2.5052323588092878e-05,
|
| 485 |
+
"loss": 0.6809,
|
| 486 |
+
"step": 340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.9249329758713136,
|
| 490 |
+
"grad_norm": 0.9566739201545715,
|
| 491 |
+
"learning_rate": 2.4886714899915415e-05,
|
| 492 |
+
"loss": 0.6185,
|
| 493 |
+
"step": 345
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.938337801608579,
|
| 497 |
+
"grad_norm": 1.1028639078140259,
|
| 498 |
+
"learning_rate": 2.4718947387263403e-05,
|
| 499 |
+
"loss": 0.6555,
|
| 500 |
+
"step": 350
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.9517426273458445,
|
| 504 |
+
"grad_norm": 1.0449227094650269,
|
| 505 |
+
"learning_rate": 2.4549057683196054e-05,
|
| 506 |
+
"loss": 0.6629,
|
| 507 |
+
"step": 355
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.9651474530831099,
|
| 511 |
+
"grad_norm": 1.186073899269104,
|
| 512 |
+
"learning_rate": 2.4377082884166016e-05,
|
| 513 |
+
"loss": 0.6665,
|
| 514 |
+
"step": 360
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.9785522788203753,
|
| 518 |
+
"grad_norm": 0.9336662292480469,
|
| 519 |
+
"learning_rate": 2.4203060541919136e-05,
|
| 520 |
+
"loss": 0.5925,
|
| 521 |
+
"step": 365
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.9919571045576407,
|
| 525 |
+
"grad_norm": 0.9751836657524109,
|
| 526 |
+
"learning_rate": 2.4027028655294804e-05,
|
| 527 |
+
"loss": 0.5957,
|
| 528 |
+
"step": 370
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.0053619302949062,
|
| 532 |
+
"grad_norm": 1.0686187744140625,
|
| 533 |
+
"learning_rate": 2.384902566192867e-05,
|
| 534 |
+
"loss": 0.5863,
|
| 535 |
+
"step": 375
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.0187667560321716,
|
| 539 |
+
"grad_norm": 1.0750106573104858,
|
| 540 |
+
"learning_rate": 2.366909042985956e-05,
|
| 541 |
+
"loss": 0.5347,
|
| 542 |
+
"step": 380
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.032171581769437,
|
| 546 |
+
"grad_norm": 1.0693711042404175,
|
| 547 |
+
"learning_rate": 2.3487262249042412e-05,
|
| 548 |
+
"loss": 0.5365,
|
| 549 |
+
"step": 385
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.0455764075067024,
|
| 553 |
+
"grad_norm": 0.9094949960708618,
|
| 554 |
+
"learning_rate": 2.330358082276905e-05,
|
| 555 |
+
"loss": 0.5336,
|
| 556 |
+
"step": 390
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.0589812332439679,
|
| 560 |
+
"grad_norm": 1.031535267829895,
|
| 561 |
+
"learning_rate": 2.3118086258998748e-05,
|
| 562 |
+
"loss": 0.5705,
|
| 563 |
+
"step": 395
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.0723860589812333,
|
| 567 |
+
"grad_norm": 0.9758612513542175,
|
| 568 |
+
"learning_rate": 2.2930819061600413e-05,
|
| 569 |
+
"loss": 0.5211,
|
| 570 |
+
"step": 400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.0857908847184987,
|
| 574 |
+
"grad_norm": 1.0757410526275635,
|
| 575 |
+
"learning_rate": 2.27418201215083e-05,
|
| 576 |
+
"loss": 0.5639,
|
| 577 |
+
"step": 405
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.0991957104557641,
|
| 581 |
+
"grad_norm": 1.033258080482483,
|
| 582 |
+
"learning_rate": 2.2551130707793243e-05,
|
| 583 |
+
"loss": 0.5209,
|
| 584 |
+
"step": 410
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.1126005361930296,
|
| 588 |
+
"grad_norm": 2.3148107528686523,
|
| 589 |
+
"learning_rate": 2.2358792458651304e-05,
|
| 590 |
+
"loss": 0.5049,
|
| 591 |
+
"step": 415
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.126005361930295,
|
| 595 |
+
"grad_norm": 1.3143815994262695,
|
| 596 |
+
"learning_rate": 2.2164847372311804e-05,
|
| 597 |
+
"loss": 0.5074,
|
| 598 |
+
"step": 420
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.1394101876675604,
|
| 602 |
+
"grad_norm": 1.0450321435928345,
|
| 603 |
+
"learning_rate": 2.1969337797866772e-05,
|
| 604 |
+
"loss": 0.5174,
|
| 605 |
+
"step": 425
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.1528150134048256,
|
| 609 |
+
"grad_norm": 1.199095606803894,
|
| 610 |
+
"learning_rate": 2.1772306426023737e-05,
|
| 611 |
+
"loss": 0.4997,
|
| 612 |
+
"step": 430
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.1662198391420913,
|
| 616 |
+
"grad_norm": 1.0282012224197388,
|
| 617 |
+
"learning_rate": 2.1573796279783987e-05,
|
| 618 |
+
"loss": 0.4671,
|
| 619 |
+
"step": 435
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.1796246648793565,
|
| 623 |
+
"grad_norm": 1.1715396642684937,
|
| 624 |
+
"learning_rate": 2.137385070504821e-05,
|
| 625 |
+
"loss": 0.5437,
|
| 626 |
+
"step": 440
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.193029490616622,
|
| 630 |
+
"grad_norm": 1.0932139158248901,
|
| 631 |
+
"learning_rate": 2.1172513361151643e-05,
|
| 632 |
+
"loss": 0.4993,
|
| 633 |
+
"step": 445
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.2064343163538873,
|
| 637 |
+
"grad_norm": 0.9794313311576843,
|
| 638 |
+
"learning_rate": 2.0969828211330824e-05,
|
| 639 |
+
"loss": 0.5142,
|
| 640 |
+
"step": 450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.2198391420911527,
|
| 644 |
+
"grad_norm": 0.9845211505889893,
|
| 645 |
+
"learning_rate": 2.0765839513123923e-05,
|
| 646 |
+
"loss": 0.4902,
|
| 647 |
+
"step": 455
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.2332439678284182,
|
| 651 |
+
"grad_norm": 1.1693155765533447,
|
| 652 |
+
"learning_rate": 2.056059180870684e-05,
|
| 653 |
+
"loss": 0.4812,
|
| 654 |
+
"step": 460
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.2466487935656836,
|
| 658 |
+
"grad_norm": 1.0942554473876953,
|
| 659 |
+
"learning_rate": 2.0354129915167175e-05,
|
| 660 |
+
"loss": 0.5094,
|
| 661 |
+
"step": 465
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.260053619302949,
|
| 665 |
+
"grad_norm": 1.0578761100769043,
|
| 666 |
+
"learning_rate": 2.014649891471811e-05,
|
| 667 |
+
"loss": 0.4843,
|
| 668 |
+
"step": 470
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.2734584450402144,
|
| 672 |
+
"grad_norm": 1.0976276397705078,
|
| 673 |
+
"learning_rate": 1.9937744144854446e-05,
|
| 674 |
+
"loss": 0.4965,
|
| 675 |
+
"step": 475
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.2868632707774799,
|
| 679 |
+
"grad_norm": 1.304211139678955,
|
| 680 |
+
"learning_rate": 1.9727911188452903e-05,
|
| 681 |
+
"loss": 0.5413,
|
| 682 |
+
"step": 480
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.3002680965147453,
|
| 686 |
+
"grad_norm": 1.3818888664245605,
|
| 687 |
+
"learning_rate": 1.9517045863818792e-05,
|
| 688 |
+
"loss": 0.5436,
|
| 689 |
+
"step": 485
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.3136729222520107,
|
| 693 |
+
"grad_norm": 1.1613248586654663,
|
| 694 |
+
"learning_rate": 1.930519421468133e-05,
|
| 695 |
+
"loss": 0.4902,
|
| 696 |
+
"step": 490
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.3270777479892761,
|
| 700 |
+
"grad_norm": 1.0429706573486328,
|
| 701 |
+
"learning_rate": 1.9092402500139693e-05,
|
| 702 |
+
"loss": 0.5013,
|
| 703 |
+
"step": 495
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.3404825737265416,
|
| 707 |
+
"grad_norm": 1.0879743099212646,
|
| 708 |
+
"learning_rate": 1.8878717184562078e-05,
|
| 709 |
+
"loss": 0.4853,
|
| 710 |
+
"step": 500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.353887399463807,
|
| 714 |
+
"grad_norm": 1.1582545042037964,
|
| 715 |
+
"learning_rate": 1.8664184927439907e-05,
|
| 716 |
+
"loss": 0.4574,
|
| 717 |
+
"step": 505
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.3672922252010724,
|
| 721 |
+
"grad_norm": 1.0746876001358032,
|
| 722 |
+
"learning_rate": 1.8448852573199465e-05,
|
| 723 |
+
"loss": 0.4626,
|
| 724 |
+
"step": 510
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.3806970509383378,
|
| 728 |
+
"grad_norm": 1.1986364126205444,
|
| 729 |
+
"learning_rate": 1.823276714097311e-05,
|
| 730 |
+
"loss": 0.4319,
|
| 731 |
+
"step": 515
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.3941018766756033,
|
| 735 |
+
"grad_norm": 1.1719796657562256,
|
| 736 |
+
"learning_rate": 1.8015975814332373e-05,
|
| 737 |
+
"loss": 0.4701,
|
| 738 |
+
"step": 520
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.4075067024128687,
|
| 742 |
+
"grad_norm": 1.3504611253738403,
|
| 743 |
+
"learning_rate": 1.779852593098513e-05,
|
| 744 |
+
"loss": 0.436,
|
| 745 |
+
"step": 525
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.420911528150134,
|
| 749 |
+
"grad_norm": 1.1986589431762695,
|
| 750 |
+
"learning_rate": 1.7580464972439137e-05,
|
| 751 |
+
"loss": 0.4929,
|
| 752 |
+
"step": 530
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.4343163538873995,
|
| 756 |
+
"grad_norm": 1.1194978952407837,
|
| 757 |
+
"learning_rate": 1.736184055363414e-05,
|
| 758 |
+
"loss": 0.4459,
|
| 759 |
+
"step": 535
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.447721179624665,
|
| 763 |
+
"grad_norm": 1.290337085723877,
|
| 764 |
+
"learning_rate": 1.7142700412544867e-05,
|
| 765 |
+
"loss": 0.4153,
|
| 766 |
+
"step": 540
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.4611260053619302,
|
| 770 |
+
"grad_norm": 0.9937898516654968,
|
| 771 |
+
"learning_rate": 1.692309239975715e-05,
|
| 772 |
+
"loss": 0.4374,
|
| 773 |
+
"step": 545
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.4745308310991958,
|
| 777 |
+
"grad_norm": 1.1232582330703735,
|
| 778 |
+
"learning_rate": 1.670306446801947e-05,
|
| 779 |
+
"loss": 0.4724,
|
| 780 |
+
"step": 550
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.487935656836461,
|
| 784 |
+
"grad_norm": 1.0209157466888428,
|
| 785 |
+
"learning_rate": 1.6482664661772153e-05,
|
| 786 |
+
"loss": 0.4321,
|
| 787 |
+
"step": 555
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.5013404825737267,
|
| 791 |
+
"grad_norm": 1.0886987447738647,
|
| 792 |
+
"learning_rate": 1.626194110665661e-05,
|
| 793 |
+
"loss": 0.4133,
|
| 794 |
+
"step": 560
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.5147453083109919,
|
| 798 |
+
"grad_norm": 1.0784316062927246,
|
| 799 |
+
"learning_rate": 1.6040941999006768e-05,
|
| 800 |
+
"loss": 0.4336,
|
| 801 |
+
"step": 565
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.5281501340482575,
|
| 805 |
+
"grad_norm": 1.1950879096984863,
|
| 806 |
+
"learning_rate": 1.581971559532515e-05,
|
| 807 |
+
"loss": 0.4528,
|
| 808 |
+
"step": 570
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.5415549597855227,
|
| 812 |
+
"grad_norm": 1.1852750778198242,
|
| 813 |
+
"learning_rate": 1.559831020174576e-05,
|
| 814 |
+
"loss": 0.4013,
|
| 815 |
+
"step": 575
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.5549597855227884,
|
| 819 |
+
"grad_norm": 1.2094613313674927,
|
| 820 |
+
"learning_rate": 1.5376774163486104e-05,
|
| 821 |
+
"loss": 0.4055,
|
| 822 |
+
"step": 580
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.5683646112600536,
|
| 826 |
+
"grad_norm": 1.06105375289917,
|
| 827 |
+
"learning_rate": 1.5155155854290777e-05,
|
| 828 |
+
"loss": 0.4311,
|
| 829 |
+
"step": 585
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.5817694369973192,
|
| 833 |
+
"grad_norm": 1.1074191331863403,
|
| 834 |
+
"learning_rate": 1.493350366586873e-05,
|
| 835 |
+
"loss": 0.449,
|
| 836 |
+
"step": 590
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.5951742627345844,
|
| 840 |
+
"grad_norm": 1.2042276859283447,
|
| 841 |
+
"learning_rate": 1.4711865997326648e-05,
|
| 842 |
+
"loss": 0.4303,
|
| 843 |
+
"step": 595
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.6085790884718498,
|
| 847 |
+
"grad_norm": 1.2283929586410522,
|
| 848 |
+
"learning_rate": 1.4490291244600712e-05,
|
| 849 |
+
"loss": 0.4116,
|
| 850 |
+
"step": 600
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.6219839142091153,
|
| 854 |
+
"grad_norm": 1.035912036895752,
|
| 855 |
+
"learning_rate": 1.4268827789889054e-05,
|
| 856 |
+
"loss": 0.4216,
|
| 857 |
+
"step": 605
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.6353887399463807,
|
| 861 |
+
"grad_norm": 1.076318383216858,
|
| 862 |
+
"learning_rate": 1.4047523991087194e-05,
|
| 863 |
+
"loss": 0.3904,
|
| 864 |
+
"step": 610
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.648793565683646,
|
| 868 |
+
"grad_norm": 1.3118699789047241,
|
| 869 |
+
"learning_rate": 1.3826428171228824e-05,
|
| 870 |
+
"loss": 0.3902,
|
| 871 |
+
"step": 615
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.6621983914209115,
|
| 875 |
+
"grad_norm": 1.1527796983718872,
|
| 876 |
+
"learning_rate": 1.3605588607934153e-05,
|
| 877 |
+
"loss": 0.3938,
|
| 878 |
+
"step": 620
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.675603217158177,
|
| 882 |
+
"grad_norm": 1.0946987867355347,
|
| 883 |
+
"learning_rate": 1.3385053522868229e-05,
|
| 884 |
+
"loss": 0.3789,
|
| 885 |
+
"step": 625
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 1.6890080428954424,
|
| 889 |
+
"grad_norm": 1.0634331703186035,
|
| 890 |
+
"learning_rate": 1.3164871071211399e-05,
|
| 891 |
+
"loss": 0.3894,
|
| 892 |
+
"step": 630
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 1.7024128686327078,
|
| 896 |
+
"grad_norm": 1.1408213376998901,
|
| 897 |
+
"learning_rate": 1.2945089331144364e-05,
|
| 898 |
+
"loss": 0.4033,
|
| 899 |
+
"step": 635
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 1.7158176943699732,
|
| 903 |
+
"grad_norm": 1.238582968711853,
|
| 904 |
+
"learning_rate": 1.2725756293350011e-05,
|
| 905 |
+
"loss": 0.4068,
|
| 906 |
+
"step": 640
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 1.7292225201072386,
|
| 910 |
+
"grad_norm": 1.1031956672668457,
|
| 911 |
+
"learning_rate": 1.2506919850534343e-05,
|
| 912 |
+
"loss": 0.3751,
|
| 913 |
+
"step": 645
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.742627345844504,
|
| 917 |
+
"grad_norm": 0.9783279299736023,
|
| 918 |
+
"learning_rate": 1.2288627786968826e-05,
|
| 919 |
+
"loss": 0.3667,
|
| 920 |
+
"step": 650
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 1.7560321715817695,
|
| 924 |
+
"grad_norm": 1.234369158744812,
|
| 925 |
+
"learning_rate": 1.2070927768056399e-05,
|
| 926 |
+
"loss": 0.3462,
|
| 927 |
+
"step": 655
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 1.7694369973190347,
|
| 931 |
+
"grad_norm": 1.101770043373108,
|
| 932 |
+
"learning_rate": 1.1853867329923436e-05,
|
| 933 |
+
"loss": 0.3976,
|
| 934 |
+
"step": 660
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 1.7828418230563003,
|
| 938 |
+
"grad_norm": 1.1363108158111572,
|
| 939 |
+
"learning_rate": 1.163749386903995e-05,
|
| 940 |
+
"loss": 0.3607,
|
| 941 |
+
"step": 665
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.7962466487935655,
|
| 945 |
+
"grad_norm": 1.2321057319641113,
|
| 946 |
+
"learning_rate": 1.1421854631870291e-05,
|
| 947 |
+
"loss": 0.384,
|
| 948 |
+
"step": 670
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 1.8096514745308312,
|
| 952 |
+
"grad_norm": 1.101366400718689,
|
| 953 |
+
"learning_rate": 1.1206996704556575e-05,
|
| 954 |
+
"loss": 0.4095,
|
| 955 |
+
"step": 675
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 1.8230563002680964,
|
| 959 |
+
"grad_norm": 1.4615360498428345,
|
| 960 |
+
"learning_rate": 1.0992967002637148e-05,
|
| 961 |
+
"loss": 0.3794,
|
| 962 |
+
"step": 680
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 1.836461126005362,
|
| 966 |
+
"grad_norm": 1.1510567665100098,
|
| 967 |
+
"learning_rate": 1.0779812260802303e-05,
|
| 968 |
+
"loss": 0.3664,
|
| 969 |
+
"step": 685
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 1.8498659517426272,
|
| 973 |
+
"grad_norm": 1.039561152458191,
|
| 974 |
+
"learning_rate": 1.056757902268945e-05,
|
| 975 |
+
"loss": 0.4016,
|
| 976 |
+
"step": 690
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 1.863270777479893,
|
| 980 |
+
"grad_norm": 1.1654466390609741,
|
| 981 |
+
"learning_rate": 1.035631363072005e-05,
|
| 982 |
+
"loss": 0.3947,
|
| 983 |
+
"step": 695
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 1.876675603217158,
|
| 987 |
+
"grad_norm": 1.1753458976745605,
|
| 988 |
+
"learning_rate": 1.014606221598046e-05,
|
| 989 |
+
"loss": 0.3328,
|
| 990 |
+
"step": 700
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 1.8900804289544237,
|
| 994 |
+
"grad_norm": 1.1596604585647583,
|
| 995 |
+
"learning_rate": 9.936870688148924e-06,
|
| 996 |
+
"loss": 0.386,
|
| 997 |
+
"step": 705
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 1.903485254691689,
|
| 1001 |
+
"grad_norm": 1.0891685485839844,
|
| 1002 |
+
"learning_rate": 9.728784725470913e-06,
|
| 1003 |
+
"loss": 0.3767,
|
| 1004 |
+
"step": 710
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.9168900804289544,
|
| 1008 |
+
"grad_norm": 1.1352084875106812,
|
| 1009 |
+
"learning_rate": 9.521849764785018e-06,
|
| 1010 |
+
"loss": 0.3346,
|
| 1011 |
+
"step": 715
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 1.9302949061662198,
|
| 1015 |
+
"grad_norm": 1.1333717107772827,
|
| 1016 |
+
"learning_rate": 9.31611099160152e-06,
|
| 1017 |
+
"loss": 0.3706,
|
| 1018 |
+
"step": 720
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 1.9436997319034852,
|
| 1022 |
+
"grad_norm": 1.1660853624343872,
|
| 1023 |
+
"learning_rate": 9.111613330235866e-06,
|
| 1024 |
+
"loss": 0.3489,
|
| 1025 |
+
"step": 725
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 1.9571045576407506,
|
| 1029 |
+
"grad_norm": 1.1579868793487549,
|
| 1030 |
+
"learning_rate": 8.90840143399917e-06,
|
| 1031 |
+
"loss": 0.3801,
|
| 1032 |
+
"step": 730
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 1.970509383378016,
|
| 1036 |
+
"grad_norm": 1.192337155342102,
|
| 1037 |
+
"learning_rate": 8.706519675447898e-06,
|
| 1038 |
+
"loss": 0.3304,
|
| 1039 |
+
"step": 735
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 1.9839142091152815,
|
| 1043 |
+
"grad_norm": 1.2775137424468994,
|
| 1044 |
+
"learning_rate": 8.506012136694832e-06,
|
| 1045 |
+
"loss": 0.3452,
|
| 1046 |
+
"step": 740
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.997319034852547,
|
| 1050 |
+
"grad_norm": 1.1246819496154785,
|
| 1051 |
+
"learning_rate": 8.306922599783491e-06,
|
| 1052 |
+
"loss": 0.362,
|
| 1053 |
+
"step": 745
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.0107238605898123,
|
| 1057 |
+
"grad_norm": 0.9721271991729736,
|
| 1058 |
+
"learning_rate": 8.109294537128057e-06,
|
| 1059 |
+
"loss": 0.3169,
|
| 1060 |
+
"step": 750
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.0241286863270775,
|
| 1064 |
+
"grad_norm": 1.1820639371871948,
|
| 1065 |
+
"learning_rate": 7.91317110202087e-06,
|
| 1066 |
+
"loss": 0.2801,
|
| 1067 |
+
"step": 755
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.037533512064343,
|
| 1071 |
+
"grad_norm": 1.1740771532058716,
|
| 1072 |
+
"learning_rate": 7.718595119209691e-06,
|
| 1073 |
+
"loss": 0.3133,
|
| 1074 |
+
"step": 760
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.0509383378016084,
|
| 1078 |
+
"grad_norm": 1.3226786851882935,
|
| 1079 |
+
"learning_rate": 7.525609075546649e-06,
|
| 1080 |
+
"loss": 0.2753,
|
| 1081 |
+
"step": 765
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.064343163538874,
|
| 1085 |
+
"grad_norm": 1.2498587369918823,
|
| 1086 |
+
"learning_rate": 7.334255110710933e-06,
|
| 1087 |
+
"loss": 0.2996,
|
| 1088 |
+
"step": 770
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.0777479892761392,
|
| 1092 |
+
"grad_norm": 1.1227176189422607,
|
| 1093 |
+
"learning_rate": 7.1445750080073964e-06,
|
| 1094 |
+
"loss": 0.2727,
|
| 1095 |
+
"step": 775
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.091152815013405,
|
| 1099 |
+
"grad_norm": 1.1149563789367676,
|
| 1100 |
+
"learning_rate": 6.956610185242891e-06,
|
| 1101 |
+
"loss": 0.2945,
|
| 1102 |
+
"step": 780
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.10455764075067,
|
| 1106 |
+
"grad_norm": 1.2285109758377075,
|
| 1107 |
+
"learning_rate": 6.770401685682417e-06,
|
| 1108 |
+
"loss": 0.2762,
|
| 1109 |
+
"step": 785
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.1179624664879357,
|
| 1113 |
+
"grad_norm": 1.1361923217773438,
|
| 1114 |
+
"learning_rate": 6.585990169087112e-06,
|
| 1115 |
+
"loss": 0.2984,
|
| 1116 |
+
"step": 790
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.131367292225201,
|
| 1120 |
+
"grad_norm": 1.0839574337005615,
|
| 1121 |
+
"learning_rate": 6.40341590283593e-06,
|
| 1122 |
+
"loss": 0.2708,
|
| 1123 |
+
"step": 795
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.1447721179624666,
|
| 1127 |
+
"grad_norm": 1.0973879098892212,
|
| 1128 |
+
"learning_rate": 6.222718753132994e-06,
|
| 1129 |
+
"loss": 0.2721,
|
| 1130 |
+
"step": 800
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.158176943699732,
|
| 1134 |
+
"grad_norm": 1.155579686164856,
|
| 1135 |
+
"learning_rate": 6.043938176302596e-06,
|
| 1136 |
+
"loss": 0.2689,
|
| 1137 |
+
"step": 805
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.1715817694369974,
|
| 1141 |
+
"grad_norm": 1.289603590965271,
|
| 1142 |
+
"learning_rate": 5.8671132101736625e-06,
|
| 1143 |
+
"loss": 0.3008,
|
| 1144 |
+
"step": 810
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.1849865951742626,
|
| 1148 |
+
"grad_norm": 1.242112159729004,
|
| 1149 |
+
"learning_rate": 5.692282465555585e-06,
|
| 1150 |
+
"loss": 0.254,
|
| 1151 |
+
"step": 815
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.1983914209115283,
|
| 1155 |
+
"grad_norm": 1.2852767705917358,
|
| 1156 |
+
"learning_rate": 5.519484117807341e-06,
|
| 1157 |
+
"loss": 0.2664,
|
| 1158 |
+
"step": 820
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.2117962466487935,
|
| 1162 |
+
"grad_norm": 1.2217613458633423,
|
| 1163 |
+
"learning_rate": 5.348755898501662e-06,
|
| 1164 |
+
"loss": 0.288,
|
| 1165 |
+
"step": 825
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.225201072386059,
|
| 1169 |
+
"grad_norm": 1.1166845560073853,
|
| 1170 |
+
"learning_rate": 5.180135087186101e-06,
|
| 1171 |
+
"loss": 0.2741,
|
| 1172 |
+
"step": 830
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.2386058981233243,
|
| 1176 |
+
"grad_norm": 1.089866280555725,
|
| 1177 |
+
"learning_rate": 5.013658503242845e-06,
|
| 1178 |
+
"loss": 0.3089,
|
| 1179 |
+
"step": 835
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.25201072386059,
|
| 1183 |
+
"grad_norm": 1.0453401803970337,
|
| 1184 |
+
"learning_rate": 4.849362497848947e-06,
|
| 1185 |
+
"loss": 0.2575,
|
| 1186 |
+
"step": 840
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.265415549597855,
|
| 1190 |
+
"grad_norm": 1.0049515962600708,
|
| 1191 |
+
"learning_rate": 4.687282946038842e-06,
|
| 1192 |
+
"loss": 0.2697,
|
| 1193 |
+
"step": 845
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.278820375335121,
|
| 1197 |
+
"grad_norm": 1.1833009719848633,
|
| 1198 |
+
"learning_rate": 4.527455238870821e-06,
|
| 1199 |
+
"loss": 0.2894,
|
| 1200 |
+
"step": 850
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.292225201072386,
|
| 1204 |
+
"grad_norm": 1.3225071430206299,
|
| 1205 |
+
"learning_rate": 4.3699142756991635e-06,
|
| 1206 |
+
"loss": 0.2672,
|
| 1207 |
+
"step": 855
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.3056300268096512,
|
| 1211 |
+
"grad_norm": 1.1755006313323975,
|
| 1212 |
+
"learning_rate": 4.2146944565536485e-06,
|
| 1213 |
+
"loss": 0.2879,
|
| 1214 |
+
"step": 860
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.319034852546917,
|
| 1218 |
+
"grad_norm": 1.1457116603851318,
|
| 1219 |
+
"learning_rate": 4.061829674628116e-06,
|
| 1220 |
+
"loss": 0.2907,
|
| 1221 |
+
"step": 865
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.3324396782841825,
|
| 1225 |
+
"grad_norm": 1.122604250907898,
|
| 1226 |
+
"learning_rate": 3.911353308879673e-06,
|
| 1227 |
+
"loss": 0.2375,
|
| 1228 |
+
"step": 870
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.3458445040214477,
|
| 1232 |
+
"grad_norm": 1.0773766040802002,
|
| 1233 |
+
"learning_rate": 3.763298216740176e-06,
|
| 1234 |
+
"loss": 0.2546,
|
| 1235 |
+
"step": 875
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.359249329758713,
|
| 1239 |
+
"grad_norm": 1.158325433731079,
|
| 1240 |
+
"learning_rate": 3.617696726941645e-06,
|
| 1241 |
+
"loss": 0.2842,
|
| 1242 |
+
"step": 880
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.3726541554959786,
|
| 1246 |
+
"grad_norm": 1.0826832056045532,
|
| 1247 |
+
"learning_rate": 3.474580632457067e-06,
|
| 1248 |
+
"loss": 0.2757,
|
| 1249 |
+
"step": 885
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.386058981233244,
|
| 1253 |
+
"grad_norm": 1.1899993419647217,
|
| 1254 |
+
"learning_rate": 3.333981183558196e-06,
|
| 1255 |
+
"loss": 0.2736,
|
| 1256 |
+
"step": 890
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.3994638069705094,
|
| 1260 |
+
"grad_norm": 1.191117525100708,
|
| 1261 |
+
"learning_rate": 3.1959290809918953e-06,
|
| 1262 |
+
"loss": 0.2469,
|
| 1263 |
+
"step": 895
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.4128686327077746,
|
| 1267 |
+
"grad_norm": 1.150436282157898,
|
| 1268 |
+
"learning_rate": 3.060454469276423e-06,
|
| 1269 |
+
"loss": 0.2686,
|
| 1270 |
+
"step": 900
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.4262734584450403,
|
| 1274 |
+
"grad_norm": 1.097445011138916,
|
| 1275 |
+
"learning_rate": 2.9275869301191855e-06,
|
| 1276 |
+
"loss": 0.2735,
|
| 1277 |
+
"step": 905
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.4396782841823055,
|
| 1281 |
+
"grad_norm": 1.221527338027954,
|
| 1282 |
+
"learning_rate": 2.7973554759574116e-06,
|
| 1283 |
+
"loss": 0.2953,
|
| 1284 |
+
"step": 910
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.453083109919571,
|
| 1288 |
+
"grad_norm": 1.2584657669067383,
|
| 1289 |
+
"learning_rate": 2.6697885436231023e-06,
|
| 1290 |
+
"loss": 0.2498,
|
| 1291 |
+
"step": 915
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.4664879356568363,
|
| 1295 |
+
"grad_norm": 1.1825857162475586,
|
| 1296 |
+
"learning_rate": 2.5449139881336587e-06,
|
| 1297 |
+
"loss": 0.2816,
|
| 1298 |
+
"step": 920
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.479892761394102,
|
| 1302 |
+
"grad_norm": 1.2222062349319458,
|
| 1303 |
+
"learning_rate": 2.422759076609597e-06,
|
| 1304 |
+
"loss": 0.2868,
|
| 1305 |
+
"step": 925
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.493297587131367,
|
| 1309 |
+
"grad_norm": 1.2197879552841187,
|
| 1310 |
+
"learning_rate": 2.303350482320592e-06,
|
| 1311 |
+
"loss": 0.2765,
|
| 1312 |
+
"step": 930
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.506702412868633,
|
| 1316 |
+
"grad_norm": 1.0478525161743164,
|
| 1317 |
+
"learning_rate": 2.1867142788611937e-06,
|
| 1318 |
+
"loss": 0.2793,
|
| 1319 |
+
"step": 935
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.520107238605898,
|
| 1323 |
+
"grad_norm": 1.177180528640747,
|
| 1324 |
+
"learning_rate": 2.0728759344575272e-06,
|
| 1325 |
+
"loss": 0.265,
|
| 1326 |
+
"step": 940
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 2.5335120643431637,
|
| 1330 |
+
"grad_norm": 1.1697869300842285,
|
| 1331 |
+
"learning_rate": 1.961860306406133e-06,
|
| 1332 |
+
"loss": 0.2607,
|
| 1333 |
+
"step": 945
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 2.546916890080429,
|
| 1337 |
+
"grad_norm": 1.2164334058761597,
|
| 1338 |
+
"learning_rate": 1.8536916356462158e-06,
|
| 1339 |
+
"loss": 0.2565,
|
| 1340 |
+
"step": 950
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 2.5603217158176945,
|
| 1344 |
+
"grad_norm": 1.2828480005264282,
|
| 1345 |
+
"learning_rate": 1.748393541466507e-06,
|
| 1346 |
+
"loss": 0.2425,
|
| 1347 |
+
"step": 955
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 2.5737265415549597,
|
| 1351 |
+
"grad_norm": 1.054573893547058,
|
| 1352 |
+
"learning_rate": 1.6459890163478391e-06,
|
| 1353 |
+
"loss": 0.2443,
|
| 1354 |
+
"step": 960
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 2.5871313672922254,
|
| 1358 |
+
"grad_norm": 0.9550372362136841,
|
| 1359 |
+
"learning_rate": 1.5465004209426053e-06,
|
| 1360 |
+
"loss": 0.2704,
|
| 1361 |
+
"step": 965
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 2.6005361930294906,
|
| 1365 |
+
"grad_norm": 1.1408909559249878,
|
| 1366 |
+
"learning_rate": 1.4499494791921563e-06,
|
| 1367 |
+
"loss": 0.2583,
|
| 1368 |
+
"step": 970
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 2.6139410187667558,
|
| 1372 |
+
"grad_norm": 1.234317660331726,
|
| 1373 |
+
"learning_rate": 1.3563572735832668e-06,
|
| 1374 |
+
"loss": 0.2557,
|
| 1375 |
+
"step": 975
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 2.6273458445040214,
|
| 1379 |
+
"grad_norm": 1.1538945436477661,
|
| 1380 |
+
"learning_rate": 1.2657442405446345e-06,
|
| 1381 |
+
"loss": 0.2638,
|
| 1382 |
+
"step": 980
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 2.640750670241287,
|
| 1386 |
+
"grad_norm": 1.0675137042999268,
|
| 1387 |
+
"learning_rate": 1.178130165984458e-06,
|
| 1388 |
+
"loss": 0.2706,
|
| 1389 |
+
"step": 985
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 2.6541554959785523,
|
| 1393 |
+
"grad_norm": 1.1528904438018799,
|
| 1394 |
+
"learning_rate": 1.093534180970074e-06,
|
| 1395 |
+
"loss": 0.2813,
|
| 1396 |
+
"step": 990
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 2.6675603217158175,
|
| 1400 |
+
"grad_norm": 0.9447672963142395,
|
| 1401 |
+
"learning_rate": 1.0119747575505695e-06,
|
| 1402 |
+
"loss": 0.2516,
|
| 1403 |
+
"step": 995
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 2.680965147453083,
|
| 1407 |
+
"grad_norm": 1.2417553663253784,
|
| 1408 |
+
"learning_rate": 9.334697047232849e-07,
|
| 1409 |
+
"loss": 0.2865,
|
| 1410 |
+
"step": 1000
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 2.6943699731903488,
|
| 1414 |
+
"grad_norm": 1.1432898044586182,
|
| 1415 |
+
"learning_rate": 8.580361645451257e-07,
|
| 1416 |
+
"loss": 0.2698,
|
| 1417 |
+
"step": 1005
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 2.707774798927614,
|
| 1421 |
+
"grad_norm": 1.1654596328735352,
|
| 1422 |
+
"learning_rate": 7.856906083894871e-07,
|
| 1423 |
+
"loss": 0.2697,
|
| 1424 |
+
"step": 1010
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 2.721179624664879,
|
| 1428 |
+
"grad_norm": 1.143183946609497,
|
| 1429 |
+
"learning_rate": 7.16448833349621e-07,
|
| 1430 |
+
"loss": 0.2382,
|
| 1431 |
+
"step": 1015
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 2.734584450402145,
|
| 1435 |
+
"grad_norm": 1.1008965969085693,
|
| 1436 |
+
"learning_rate": 6.503259587892535e-07,
|
| 1437 |
+
"loss": 0.2502,
|
| 1438 |
+
"step": 1020
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 2.7479892761394105,
|
| 1442 |
+
"grad_norm": 1.171980857849121,
|
| 1443 |
+
"learning_rate": 5.873364230411754e-07,
|
| 1444 |
+
"loss": 0.2462,
|
| 1445 |
+
"step": 1025
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 2.7613941018766757,
|
| 1449 |
+
"grad_norm": 1.036729097366333,
|
| 1450 |
+
"learning_rate": 5.274939802545415e-07,
|
| 1451 |
+
"loss": 0.2596,
|
| 1452 |
+
"step": 1030
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.774798927613941,
|
| 1456 |
+
"grad_norm": 1.177030324935913,
|
| 1457 |
+
"learning_rate": 4.7081169739157717e-07,
|
| 1458 |
+
"loss": 0.2802,
|
| 1459 |
+
"step": 1035
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 2.7882037533512065,
|
| 1463 |
+
"grad_norm": 1.1302504539489746,
|
| 1464 |
+
"learning_rate": 4.173019513743198e-07,
|
| 1465 |
+
"loss": 0.2842,
|
| 1466 |
+
"step": 1040
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 2.8016085790884717,
|
| 1470 |
+
"grad_norm": 1.1969364881515503,
|
| 1471 |
+
"learning_rate": 3.6697642638204354e-07,
|
| 1472 |
+
"loss": 0.2594,
|
| 1473 |
+
"step": 1045
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 2.8150134048257374,
|
| 1477 |
+
"grad_norm": 1.1496009826660156,
|
| 1478 |
+
"learning_rate": 3.198461112999468e-07,
|
| 1479 |
+
"loss": 0.283,
|
| 1480 |
+
"step": 1050
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 2.8284182305630026,
|
| 1484 |
+
"grad_norm": 1.0756111145019531,
|
| 1485 |
+
"learning_rate": 2.7592129731967176e-07,
|
| 1486 |
+
"loss": 0.2784,
|
| 1487 |
+
"step": 1055
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 2.841823056300268,
|
| 1491 |
+
"grad_norm": 1.0963841676712036,
|
| 1492 |
+
"learning_rate": 2.3521157569214024e-07,
|
| 1493 |
+
"loss": 0.2681,
|
| 1494 |
+
"step": 1060
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 2.8552278820375334,
|
| 1498 |
+
"grad_norm": 1.1973907947540283,
|
| 1499 |
+
"learning_rate": 1.9772583563326729e-07,
|
| 1500 |
+
"loss": 0.2839,
|
| 1501 |
+
"step": 1065
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 2.868632707774799,
|
| 1505 |
+
"grad_norm": 1.1479398012161255,
|
| 1506 |
+
"learning_rate": 1.6347226238293377e-07,
|
| 1507 |
+
"loss": 0.2454,
|
| 1508 |
+
"step": 1070
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 2.8820375335120643,
|
| 1512 |
+
"grad_norm": 1.0420469045639038,
|
| 1513 |
+
"learning_rate": 1.3245833541768947e-07,
|
| 1514 |
+
"loss": 0.2409,
|
| 1515 |
+
"step": 1075
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 2.89544235924933,
|
| 1519 |
+
"grad_norm": 1.1466476917266846,
|
| 1520 |
+
"learning_rate": 1.0469082681757114e-07,
|
| 1521 |
+
"loss": 0.263,
|
| 1522 |
+
"step": 1080
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 2.908847184986595,
|
| 1526 |
+
"grad_norm": 1.198553442955017,
|
| 1527 |
+
"learning_rate": 8.0175799787372e-08,
|
| 1528 |
+
"loss": 0.2287,
|
| 1529 |
+
"step": 1085
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 2.9222520107238603,
|
| 1533 |
+
"grad_norm": 1.0365747213363647,
|
| 1534 |
+
"learning_rate": 5.891860733270249e-08,
|
| 1535 |
+
"loss": 0.2552,
|
| 1536 |
+
"step": 1090
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 2.935656836461126,
|
| 1540 |
+
"grad_norm": 1.1062986850738525,
|
| 1541 |
+
"learning_rate": 4.092389109113526e-08,
|
| 1542 |
+
"loss": 0.246,
|
| 1543 |
+
"step": 1095
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 2.9490616621983916,
|
| 1547 |
+
"grad_norm": 1.0808930397033691,
|
| 1548 |
+
"learning_rate": 2.6195580318667533e-08,
|
| 1549 |
+
"loss": 0.2531,
|
| 1550 |
+
"step": 1100
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 2.962466487935657,
|
| 1554 |
+
"grad_norm": 1.215558648109436,
|
| 1555 |
+
"learning_rate": 1.4736891031752464e-08,
|
| 1556 |
+
"loss": 0.254,
|
| 1557 |
+
"step": 1105
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 2.975871313672922,
|
| 1561 |
+
"grad_norm": 1.107862114906311,
|
| 1562 |
+
"learning_rate": 6.550325305054172e-09,
|
| 1563 |
+
"loss": 0.2557,
|
| 1564 |
+
"step": 1110
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 2.9892761394101877,
|
| 1568 |
+
"grad_norm": 1.1284103393554688,
|
| 1569 |
+
"learning_rate": 1.6376707251097855e-09,
|
| 1570 |
+
"loss": 0.247,
|
| 1571 |
+
"step": 1115
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 3.0,
|
| 1575 |
+
"step": 1119,
|
| 1576 |
+
"total_flos": 1.647805420300927e+18,
|
| 1577 |
+
"train_loss": 0.5340931572159877,
|
| 1578 |
+
"train_runtime": 781.1418,
|
| 1579 |
+
"train_samples_per_second": 45.821,
|
| 1580 |
+
"train_steps_per_second": 1.433
|
| 1581 |
+
}
|
| 1582 |
+
],
|
| 1583 |
+
"logging_steps": 5,
|
| 1584 |
+
"max_steps": 1119,
|
| 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": 1.647805420300927e+18,
|
| 1601 |
+
"train_batch_size": 2,
|
| 1602 |
+
"trial_name": null,
|
| 1603 |
+
"trial_params": null
|
| 1604 |
+
}
|
4_128_e3_3e-5/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eedd797c52a048cbb82c620bc0f76da294abbf5058459a6987bdc745abaf434d
|
| 3 |
+
size 8145
|
4_128_e3_3e-5/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|