Upload folder using huggingface_hub
Browse files- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/README.md +82 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/added_tokens.json +28 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/all_results.json +13 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/config.json +45 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/eval_results.json +8 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/generation_config.json +7 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/merges.txt +0 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/model.safetensors +3 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/special_tokens_map.json +31 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/tokenizer_config.json +241 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/train_results.json +8 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/trainer_log.jsonl +176 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/trainer_state.json +1283 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/training_args.bin +3 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/training_eval_loss.png +0 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/training_loss.png +0 -0
- qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/vocab.json +0 -0
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/README.md
ADDED
|
@@ -0,0 +1,82 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: other
|
| 4 |
+
base_model: /data/jrlin/HyperScale/.cache/model_specs/qwen3/qwen3/base/std/p223m
|
| 5 |
+
tags:
|
| 6 |
+
- llama-factory
|
| 7 |
+
- full
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
datasets:
|
| 10 |
+
- arrow
|
| 11 |
+
model-index:
|
| 12 |
+
- name: attempt3_20260310_211619
|
| 13 |
+
results: []
|
| 14 |
+
---
|
| 15 |
+
|
| 16 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 17 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 18 |
+
|
| 19 |
+
# attempt3_20260310_211619
|
| 20 |
+
|
| 21 |
+
This model is a fine-tuned version of [/data/jrlin/HyperScale/.cache/model_specs/qwen3/qwen3/base/std/p223m](https://huggingface.co//data/jrlin/HyperScale/.cache/model_specs/qwen3/qwen3/base/std/p223m) on the owt_local_train dataset.
|
| 22 |
+
It achieves the following results on the evaluation set:
|
| 23 |
+
- Loss: 3.5179
|
| 24 |
+
|
| 25 |
+
## Model description
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Intended uses & limitations
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training and evaluation data
|
| 34 |
+
|
| 35 |
+
More information needed
|
| 36 |
+
|
| 37 |
+
## Training procedure
|
| 38 |
+
|
| 39 |
+
### Training hyperparameters
|
| 40 |
+
|
| 41 |
+
The following hyperparameters were used during training:
|
| 42 |
+
- learning_rate: 0.0001
|
| 43 |
+
- train_batch_size: 4
|
| 44 |
+
- eval_batch_size: 8
|
| 45 |
+
- seed: 42
|
| 46 |
+
- distributed_type: multi-GPU
|
| 47 |
+
- num_devices: 2
|
| 48 |
+
- gradient_accumulation_steps: 16
|
| 49 |
+
- total_train_batch_size: 128
|
| 50 |
+
- total_eval_batch_size: 16
|
| 51 |
+
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 52 |
+
- lr_scheduler_type: cosine
|
| 53 |
+
- lr_scheduler_warmup_ratio: 0.02
|
| 54 |
+
- training_steps: 16011
|
| 55 |
+
|
| 56 |
+
### Training results
|
| 57 |
+
|
| 58 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 59 |
+
|:-------------:|:------:|:-----:|:---------------:|
|
| 60 |
+
| 5.0361 | 0.0308 | 1024 | 4.9493 |
|
| 61 |
+
| 4.3616 | 0.0616 | 2048 | 4.3255 |
|
| 62 |
+
| 4.0822 | 0.0924 | 3072 | 4.0634 |
|
| 63 |
+
| 3.9257 | 0.1232 | 4096 | 3.9077 |
|
| 64 |
+
| 3.8074 | 0.1540 | 5120 | 3.8044 |
|
| 65 |
+
| 3.7379 | 0.1848 | 6144 | 3.7293 |
|
| 66 |
+
| 3.6719 | 0.2156 | 7168 | 3.6740 |
|
| 67 |
+
| 3.6267 | 0.2464 | 8192 | 3.6296 |
|
| 68 |
+
| 3.5901 | 0.2773 | 9216 | 3.5964 |
|
| 69 |
+
| 3.5696 | 0.3081 | 10240 | 3.5699 |
|
| 70 |
+
| 3.5561 | 0.3389 | 11264 | 3.5500 |
|
| 71 |
+
| 3.5263 | 0.3697 | 12288 | 3.5356 |
|
| 72 |
+
| 3.5247 | 0.4005 | 13312 | 3.5258 |
|
| 73 |
+
| 3.521 | 0.4313 | 14336 | 3.5202 |
|
| 74 |
+
| 3.5164 | 0.4621 | 15360 | 3.5181 |
|
| 75 |
+
|
| 76 |
+
|
| 77 |
+
### Framework versions
|
| 78 |
+
|
| 79 |
+
- Transformers 4.51.3
|
| 80 |
+
- Pytorch 2.5.1+cu124
|
| 81 |
+
- Datasets 3.6.0
|
| 82 |
+
- Tokenizers 0.21.4
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/added_tokens.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</think>": 151668,
|
| 3 |
+
"</tool_call>": 151658,
|
| 4 |
+
"</tool_response>": 151666,
|
| 5 |
+
"<think>": 151667,
|
| 6 |
+
"<tool_call>": 151657,
|
| 7 |
+
"<tool_response>": 151665,
|
| 8 |
+
"<|box_end|>": 151649,
|
| 9 |
+
"<|box_start|>": 151648,
|
| 10 |
+
"<|endoftext|>": 151643,
|
| 11 |
+
"<|file_sep|>": 151664,
|
| 12 |
+
"<|fim_middle|>": 151660,
|
| 13 |
+
"<|fim_pad|>": 151662,
|
| 14 |
+
"<|fim_prefix|>": 151659,
|
| 15 |
+
"<|fim_suffix|>": 151661,
|
| 16 |
+
"<|im_end|>": 151645,
|
| 17 |
+
"<|im_start|>": 151644,
|
| 18 |
+
"<|image_pad|>": 151655,
|
| 19 |
+
"<|object_ref_end|>": 151647,
|
| 20 |
+
"<|object_ref_start|>": 151646,
|
| 21 |
+
"<|quad_end|>": 151651,
|
| 22 |
+
"<|quad_start|>": 151650,
|
| 23 |
+
"<|repo_name|>": 151663,
|
| 24 |
+
"<|video_pad|>": 151656,
|
| 25 |
+
"<|vision_end|>": 151653,
|
| 26 |
+
"<|vision_pad|>": 151654,
|
| 27 |
+
"<|vision_start|>": 151652
|
| 28 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.48167145500218106,
|
| 3 |
+
"eval_loss": 3.5179009437561035,
|
| 4 |
+
"eval_perplexity": 33.71358742401022,
|
| 5 |
+
"eval_runtime": 17.3329,
|
| 6 |
+
"eval_samples_per_second": 244.507,
|
| 7 |
+
"eval_steps_per_second": 15.289,
|
| 8 |
+
"total_flos": 2.6742109246125507e+18,
|
| 9 |
+
"train_loss": 3.9105984272824914,
|
| 10 |
+
"train_runtime": 32974.1722,
|
| 11 |
+
"train_samples_per_second": 62.152,
|
| 12 |
+
"train_steps_per_second": 0.486
|
| 13 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"dtype": "bfloat16",
|
| 9 |
+
"eos_token_id": 151645,
|
| 10 |
+
"head_dim": 128,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 2304,
|
| 15 |
+
"layer_types": [
|
| 16 |
+
"full_attention",
|
| 17 |
+
"full_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"full_attention",
|
| 20 |
+
"full_attention",
|
| 21 |
+
"full_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"full_attention",
|
| 24 |
+
"full_attention",
|
| 25 |
+
"full_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"full_attention"
|
| 28 |
+
],
|
| 29 |
+
"max_position_embeddings": 40960,
|
| 30 |
+
"max_window_layers": 12,
|
| 31 |
+
"model_type": "qwen3",
|
| 32 |
+
"num_attention_heads": 12,
|
| 33 |
+
"num_hidden_layers": 12,
|
| 34 |
+
"num_key_value_heads": 6,
|
| 35 |
+
"rms_norm_eps": 1e-06,
|
| 36 |
+
"rope_scaling": null,
|
| 37 |
+
"rope_theta": 1000000,
|
| 38 |
+
"sliding_window": null,
|
| 39 |
+
"tie_word_embeddings": true,
|
| 40 |
+
"torch_dtype": "bfloat16",
|
| 41 |
+
"transformers_version": "4.51.3",
|
| 42 |
+
"use_cache": false,
|
| 43 |
+
"use_sliding_window": false,
|
| 44 |
+
"vocab_size": 151936
|
| 45 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.48167145500218106,
|
| 3 |
+
"eval_loss": 3.5179009437561035,
|
| 4 |
+
"eval_perplexity": 33.71358742401022,
|
| 5 |
+
"eval_runtime": 17.3329,
|
| 6 |
+
"eval_samples_per_second": 244.507,
|
| 7 |
+
"eval_steps_per_second": 15.289
|
| 8 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 151643,
|
| 4 |
+
"eos_token_id": 151645,
|
| 5 |
+
"transformers_version": "4.51.3",
|
| 6 |
+
"use_cache": false
|
| 7 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a584d25e00a1411e91d52d110ae941f7ca7bc0c5589778b2c2a1f5c18bc3f353
|
| 3 |
+
size 679143656
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/special_tokens_map.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>",
|
| 5 |
+
"<|object_ref_start|>",
|
| 6 |
+
"<|object_ref_end|>",
|
| 7 |
+
"<|box_start|>",
|
| 8 |
+
"<|box_end|>",
|
| 9 |
+
"<|quad_start|>",
|
| 10 |
+
"<|quad_end|>",
|
| 11 |
+
"<|vision_start|>",
|
| 12 |
+
"<|vision_end|>",
|
| 13 |
+
"<|vision_pad|>",
|
| 14 |
+
"<|image_pad|>",
|
| 15 |
+
"<|video_pad|>"
|
| 16 |
+
],
|
| 17 |
+
"eos_token": {
|
| 18 |
+
"content": "<|im_end|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
},
|
| 24 |
+
"pad_token": {
|
| 25 |
+
"content": "<|endoftext|>",
|
| 26 |
+
"lstrip": false,
|
| 27 |
+
"normalized": false,
|
| 28 |
+
"rstrip": false,
|
| 29 |
+
"single_word": false
|
| 30 |
+
}
|
| 31 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/tokenizer_config.json
ADDED
|
@@ -0,0 +1,241 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"151643": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"151644": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"151645": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"151646": {
|
| 30 |
+
"content": "<|object_ref_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"151647": {
|
| 38 |
+
"content": "<|object_ref_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"151648": {
|
| 46 |
+
"content": "<|box_start|>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"151649": {
|
| 54 |
+
"content": "<|box_end|>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"151650": {
|
| 62 |
+
"content": "<|quad_start|>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"151651": {
|
| 70 |
+
"content": "<|quad_end|>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"151652": {
|
| 78 |
+
"content": "<|vision_start|>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"151653": {
|
| 86 |
+
"content": "<|vision_end|>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"151654": {
|
| 94 |
+
"content": "<|vision_pad|>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"151655": {
|
| 102 |
+
"content": "<|image_pad|>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"151656": {
|
| 110 |
+
"content": "<|video_pad|>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"151657": {
|
| 118 |
+
"content": "<tool_call>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": false
|
| 124 |
+
},
|
| 125 |
+
"151658": {
|
| 126 |
+
"content": "</tool_call>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": false
|
| 132 |
+
},
|
| 133 |
+
"151659": {
|
| 134 |
+
"content": "<|fim_prefix|>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": false
|
| 140 |
+
},
|
| 141 |
+
"151660": {
|
| 142 |
+
"content": "<|fim_middle|>",
|
| 143 |
+
"lstrip": false,
|
| 144 |
+
"normalized": false,
|
| 145 |
+
"rstrip": false,
|
| 146 |
+
"single_word": false,
|
| 147 |
+
"special": false
|
| 148 |
+
},
|
| 149 |
+
"151661": {
|
| 150 |
+
"content": "<|fim_suffix|>",
|
| 151 |
+
"lstrip": false,
|
| 152 |
+
"normalized": false,
|
| 153 |
+
"rstrip": false,
|
| 154 |
+
"single_word": false,
|
| 155 |
+
"special": false
|
| 156 |
+
},
|
| 157 |
+
"151662": {
|
| 158 |
+
"content": "<|fim_pad|>",
|
| 159 |
+
"lstrip": false,
|
| 160 |
+
"normalized": false,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false,
|
| 163 |
+
"special": false
|
| 164 |
+
},
|
| 165 |
+
"151663": {
|
| 166 |
+
"content": "<|repo_name|>",
|
| 167 |
+
"lstrip": false,
|
| 168 |
+
"normalized": false,
|
| 169 |
+
"rstrip": false,
|
| 170 |
+
"single_word": false,
|
| 171 |
+
"special": false
|
| 172 |
+
},
|
| 173 |
+
"151664": {
|
| 174 |
+
"content": "<|file_sep|>",
|
| 175 |
+
"lstrip": false,
|
| 176 |
+
"normalized": false,
|
| 177 |
+
"rstrip": false,
|
| 178 |
+
"single_word": false,
|
| 179 |
+
"special": false
|
| 180 |
+
},
|
| 181 |
+
"151665": {
|
| 182 |
+
"content": "<tool_response>",
|
| 183 |
+
"lstrip": false,
|
| 184 |
+
"normalized": false,
|
| 185 |
+
"rstrip": false,
|
| 186 |
+
"single_word": false,
|
| 187 |
+
"special": false
|
| 188 |
+
},
|
| 189 |
+
"151666": {
|
| 190 |
+
"content": "</tool_response>",
|
| 191 |
+
"lstrip": false,
|
| 192 |
+
"normalized": false,
|
| 193 |
+
"rstrip": false,
|
| 194 |
+
"single_word": false,
|
| 195 |
+
"special": false
|
| 196 |
+
},
|
| 197 |
+
"151667": {
|
| 198 |
+
"content": "<think>",
|
| 199 |
+
"lstrip": false,
|
| 200 |
+
"normalized": false,
|
| 201 |
+
"rstrip": false,
|
| 202 |
+
"single_word": false,
|
| 203 |
+
"special": false
|
| 204 |
+
},
|
| 205 |
+
"151668": {
|
| 206 |
+
"content": "</think>",
|
| 207 |
+
"lstrip": false,
|
| 208 |
+
"normalized": false,
|
| 209 |
+
"rstrip": false,
|
| 210 |
+
"single_word": false,
|
| 211 |
+
"special": false
|
| 212 |
+
}
|
| 213 |
+
},
|
| 214 |
+
"additional_special_tokens": [
|
| 215 |
+
"<|im_start|>",
|
| 216 |
+
"<|im_end|>",
|
| 217 |
+
"<|object_ref_start|>",
|
| 218 |
+
"<|object_ref_end|>",
|
| 219 |
+
"<|box_start|>",
|
| 220 |
+
"<|box_end|>",
|
| 221 |
+
"<|quad_start|>",
|
| 222 |
+
"<|quad_end|>",
|
| 223 |
+
"<|vision_start|>",
|
| 224 |
+
"<|vision_end|>",
|
| 225 |
+
"<|vision_pad|>",
|
| 226 |
+
"<|image_pad|>",
|
| 227 |
+
"<|video_pad|>"
|
| 228 |
+
],
|
| 229 |
+
"bos_token": null,
|
| 230 |
+
"chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0].role == 'system' %}\n {{- messages[0].content + '\\n\\n' }}\n {%- endif %}\n {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0].role == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n {%- set index = (messages|length - 1) - loop.index0 %}\n {%- if ns.multi_step_tool and message.role == \"user\" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n {%- set ns.multi_step_tool = false %}\n {%- set ns.last_query_index = index %}\n {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n {%- if message.content is string %}\n {%- set content = message.content %}\n {%- else %}\n {%- set content = '' %}\n {%- endif %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n {{- '<|im_start|>' + message.role + '\\n' + content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {%- set reasoning_content = '' %}\n {%- if message.reasoning_content is string %}\n {%- set reasoning_content = message.reasoning_content %}\n {%- else %}\n {%- if '</think>' in content %}\n {%- set reasoning_content = content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n {%- set content = content.split('</think>')[-1].lstrip('\\n') %}\n {%- endif %}\n {%- endif %}\n {%- if loop.index0 > ns.last_query_index %}\n {%- if loop.last or (not loop.last and reasoning_content) %}\n {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- if message.tool_calls %}\n {%- for tool_call in message.tool_calls %}\n {%- if (loop.first and content) or (not loop.first) %}\n {{- '\\n' }}\n {%- endif %}\n {%- if tool_call.function %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {%- if tool_call.arguments is string %}\n {{- tool_call.arguments }}\n {%- else %}\n {{- tool_call.arguments | tojson }}\n {%- endif %}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {%- endif %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n {%- if enable_thinking is defined and enable_thinking is false %}\n {{- '<think>\\n\\n</think>\\n\\n' }}\n {%- endif %}\n{%- endif %}",
|
| 231 |
+
"clean_up_tokenization_spaces": false,
|
| 232 |
+
"eos_token": "<|im_end|>",
|
| 233 |
+
"errors": "replace",
|
| 234 |
+
"extra_special_tokens": {},
|
| 235 |
+
"model_max_length": 131072,
|
| 236 |
+
"pad_token": "<|endoftext|>",
|
| 237 |
+
"padding_side": "right",
|
| 238 |
+
"split_special_tokens": false,
|
| 239 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 240 |
+
"unk_token": null
|
| 241 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.48167145500218106,
|
| 3 |
+
"total_flos": 2.6742109246125507e+18,
|
| 4 |
+
"train_loss": 3.9105984272824914,
|
| 5 |
+
"train_runtime": 32974.1722,
|
| 6 |
+
"train_samples_per_second": 62.152,
|
| 7 |
+
"train_steps_per_second": 0.486
|
| 8 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/trainer_log.jsonl
ADDED
|
@@ -0,0 +1,176 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"current_steps": 100, "total_steps": 16011, "loss": 10.6918, "lr": 3.08411214953271e-05, "epoch": 0.0030083783336592408, "percentage": 0.62, "elapsed_time": "0:03:24", "remaining_time": "9:02:04"}
|
| 2 |
+
{"current_steps": 200, "total_steps": 16011, "loss": 7.9189, "lr": 6.199376947040498e-05, "epoch": 0.0060167566673184816, "percentage": 1.25, "elapsed_time": "0:06:48", "remaining_time": "8:58:08"}
|
| 3 |
+
{"current_steps": 300, "total_steps": 16011, "loss": 6.7175, "lr": 9.314641744548286e-05, "epoch": 0.009025135000977723, "percentage": 1.87, "elapsed_time": "0:10:12", "remaining_time": "8:54:25"}
|
| 4 |
+
{"current_steps": 400, "total_steps": 16011, "loss": 6.1791, "lr": 9.999390218499588e-05, "epoch": 0.012033513334636963, "percentage": 2.5, "elapsed_time": "0:13:36", "remaining_time": "8:51:16"}
|
| 5 |
+
{"current_steps": 500, "total_steps": 16011, "loss": 5.8595, "lr": 9.996824677076817e-05, "epoch": 0.015041891668296205, "percentage": 3.12, "elapsed_time": "0:17:01", "remaining_time": "8:47:59"}
|
| 6 |
+
{"current_steps": 600, "total_steps": 16011, "loss": 5.637, "lr": 9.992255893455041e-05, "epoch": 0.018050270001955446, "percentage": 3.75, "elapsed_time": "0:20:25", "remaining_time": "8:44:30"}
|
| 7 |
+
{"current_steps": 700, "total_steps": 16011, "loss": 5.4491, "lr": 9.9856856992735e-05, "epoch": 0.021058648335614688, "percentage": 4.37, "elapsed_time": "0:23:49", "remaining_time": "8:41:03"}
|
| 8 |
+
{"current_steps": 800, "total_steps": 16011, "loss": 5.3055, "lr": 9.977116728543007e-05, "epoch": 0.024067026669273926, "percentage": 5.0, "elapsed_time": "0:27:13", "remaining_time": "8:37:38"}
|
| 9 |
+
{"current_steps": 900, "total_steps": 16011, "loss": 5.1579, "lr": 9.966552416589973e-05, "epoch": 0.027075405002933168, "percentage": 5.62, "elapsed_time": "0:30:37", "remaining_time": "8:34:17"}
|
| 10 |
+
{"current_steps": 1000, "total_steps": 16011, "loss": 5.0361, "lr": 9.953996998679162e-05, "epoch": 0.03008378333659241, "percentage": 6.25, "elapsed_time": "0:34:02", "remaining_time": "8:30:53"}
|
| 11 |
+
{"current_steps": 1024, "total_steps": 16011, "eval_loss": 4.949322700500488, "epoch": 0.030805794136670627, "percentage": 6.4, "elapsed_time": "0:35:08", "remaining_time": "8:34:20"}
|
| 12 |
+
{"current_steps": 1100, "total_steps": 16011, "loss": 4.9213, "lr": 9.939455508315775e-05, "epoch": 0.03309216167025165, "percentage": 6.87, "elapsed_time": "0:37:43", "remaining_time": "8:31:19"}
|
| 13 |
+
{"current_steps": 1200, "total_steps": 16011, "loss": 4.8299, "lr": 9.92293377522749e-05, "epoch": 0.03610054000391089, "percentage": 7.49, "elapsed_time": "0:41:07", "remaining_time": "8:27:33"}
|
| 14 |
+
{"current_steps": 1300, "total_steps": 16011, "loss": 4.7423, "lr": 9.904438423027316e-05, "epoch": 0.03910891833757013, "percentage": 8.12, "elapsed_time": "0:44:31", "remaining_time": "8:23:48"}
|
| 15 |
+
{"current_steps": 1400, "total_steps": 16011, "loss": 4.662, "lr": 9.883976866558158e-05, "epoch": 0.042117296671229376, "percentage": 8.74, "elapsed_time": "0:47:55", "remaining_time": "8:20:06"}
|
| 16 |
+
{"current_steps": 1500, "total_steps": 16011, "loss": 4.5986, "lr": 9.86155730892019e-05, "epoch": 0.045125675004888614, "percentage": 9.37, "elapsed_time": "0:51:19", "remaining_time": "8:16:26"}
|
| 17 |
+
{"current_steps": 1600, "total_steps": 16011, "loss": 4.5433, "lr": 9.837188738182205e-05, "epoch": 0.04813405333854785, "percentage": 9.99, "elapsed_time": "0:54:42", "remaining_time": "8:12:48"}
|
| 18 |
+
{"current_steps": 1700, "total_steps": 16011, "loss": 4.4899, "lr": 9.81088092377827e-05, "epoch": 0.0511424316722071, "percentage": 10.62, "elapsed_time": "0:58:06", "remaining_time": "8:09:12"}
|
| 19 |
+
{"current_steps": 1800, "total_steps": 16011, "loss": 4.443, "lr": 9.782644412591135e-05, "epoch": 0.054150810005866336, "percentage": 11.24, "elapsed_time": "1:01:29", "remaining_time": "8:05:32"}
|
| 20 |
+
{"current_steps": 1900, "total_steps": 16011, "loss": 4.3915, "lr": 9.752490524723946e-05, "epoch": 0.05715918833952558, "percentage": 11.87, "elapsed_time": "1:04:53", "remaining_time": "8:01:57"}
|
| 21 |
+
{"current_steps": 2000, "total_steps": 16011, "loss": 4.3616, "lr": 9.720431348961991e-05, "epoch": 0.06016756667318482, "percentage": 12.49, "elapsed_time": "1:08:18", "remaining_time": "7:58:29"}
|
| 22 |
+
{"current_steps": 2048, "total_steps": 16011, "eval_loss": 4.325469493865967, "epoch": 0.061611588273341254, "percentage": 12.79, "elapsed_time": "1:10:13", "remaining_time": "7:58:47"}
|
| 23 |
+
{"current_steps": 2100, "total_steps": 16011, "loss": 4.3253, "lr": 9.686479737926266e-05, "epoch": 0.06317594500684406, "percentage": 13.12, "elapsed_time": "1:12:02", "remaining_time": "7:57:16"}
|
| 24 |
+
{"current_steps": 2200, "total_steps": 16011, "loss": 4.2853, "lr": 9.650649302920819e-05, "epoch": 0.0661843233405033, "percentage": 13.74, "elapsed_time": "1:15:26", "remaining_time": "7:53:34"}
|
| 25 |
+
{"current_steps": 2300, "total_steps": 16011, "loss": 4.261, "lr": 9.612954408475934e-05, "epoch": 0.06919270167416254, "percentage": 14.37, "elapsed_time": "1:18:50", "remaining_time": "7:49:57"}
|
| 26 |
+
{"current_steps": 2400, "total_steps": 16011, "loss": 4.229, "lr": 9.573410166589356e-05, "epoch": 0.07220108000782179, "percentage": 14.99, "elapsed_time": "1:22:13", "remaining_time": "7:46:20"}
|
| 27 |
+
{"current_steps": 2500, "total_steps": 16011, "loss": 4.2043, "lr": 9.532032430667834e-05, "epoch": 0.07520945834148103, "percentage": 15.61, "elapsed_time": "1:25:37", "remaining_time": "7:42:45"}
|
| 28 |
+
{"current_steps": 2600, "total_steps": 16011, "loss": 4.173, "lr": 9.488837789171455e-05, "epoch": 0.07821783667514026, "percentage": 16.24, "elapsed_time": "1:29:01", "remaining_time": "7:39:10"}
|
| 29 |
+
{"current_steps": 2700, "total_steps": 16011, "loss": 4.1537, "lr": 9.443843558963266e-05, "epoch": 0.08122621500879951, "percentage": 16.86, "elapsed_time": "1:32:25", "remaining_time": "7:35:37"}
|
| 30 |
+
{"current_steps": 2800, "total_steps": 16011, "loss": 4.1276, "lr": 9.397067778366902e-05, "epoch": 0.08423459334245875, "percentage": 17.49, "elapsed_time": "1:35:49", "remaining_time": "7:32:05"}
|
| 31 |
+
{"current_steps": 2900, "total_steps": 16011, "loss": 4.1065, "lr": 9.348529199934959e-05, "epoch": 0.08724297167611798, "percentage": 18.11, "elapsed_time": "1:39:13", "remaining_time": "7:28:35"}
|
| 32 |
+
{"current_steps": 3000, "total_steps": 16011, "loss": 4.0822, "lr": 9.298247282931048e-05, "epoch": 0.09025135000977723, "percentage": 18.74, "elapsed_time": "1:42:36", "remaining_time": "7:25:02"}
|
| 33 |
+
{"current_steps": 3072, "total_steps": 16011, "eval_loss": 4.063363075256348, "epoch": 0.09241738241001188, "percentage": 19.19, "elapsed_time": "1:45:20", "remaining_time": "7:23:42"}
|
| 34 |
+
{"current_steps": 3100, "total_steps": 16011, "loss": 4.071, "lr": 9.246242185528508e-05, "epoch": 0.09325972834343647, "percentage": 19.36, "elapsed_time": "1:46:18", "remaining_time": "7:22:43"}
|
| 35 |
+
{"current_steps": 3200, "total_steps": 16011, "loss": 4.0473, "lr": 9.192534756728936e-05, "epoch": 0.0962681066770957, "percentage": 19.99, "elapsed_time": "1:49:42", "remaining_time": "7:19:11"}
|
| 36 |
+
{"current_steps": 3300, "total_steps": 16011, "loss": 4.0297, "lr": 9.137146528003762e-05, "epoch": 0.09927648501075495, "percentage": 20.61, "elapsed_time": "1:53:06", "remaining_time": "7:15:40"}
|
| 37 |
+
{"current_steps": 3400, "total_steps": 16011, "loss": 4.0133, "lr": 9.080099704662214e-05, "epoch": 0.1022848633444142, "percentage": 21.24, "elapsed_time": "1:56:30", "remaining_time": "7:12:09"}
|
| 38 |
+
{"current_steps": 3500, "total_steps": 16011, "loss": 3.9874, "lr": 9.021417156949143e-05, "epoch": 0.10529324167807344, "percentage": 21.86, "elapsed_time": "1:59:54", "remaining_time": "7:08:38"}
|
| 39 |
+
{"current_steps": 3600, "total_steps": 16011, "loss": 3.9776, "lr": 8.961122410876268e-05, "epoch": 0.10830162001173267, "percentage": 22.48, "elapsed_time": "2:03:19", "remaining_time": "7:05:08"}
|
| 40 |
+
{"current_steps": 3700, "total_steps": 16011, "loss": 3.9607, "lr": 8.899239638790535e-05, "epoch": 0.11130999834539192, "percentage": 23.11, "elapsed_time": "2:06:43", "remaining_time": "7:01:37"}
|
| 41 |
+
{"current_steps": 3800, "total_steps": 16011, "loss": 3.9548, "lr": 8.835793649683342e-05, "epoch": 0.11431837667905116, "percentage": 23.73, "elapsed_time": "2:10:07", "remaining_time": "6:58:08"}
|
| 42 |
+
{"current_steps": 3900, "total_steps": 16011, "loss": 3.9344, "lr": 8.770809879244541e-05, "epoch": 0.1173267550127104, "percentage": 24.36, "elapsed_time": "2:13:31", "remaining_time": "6:54:39"}
|
| 43 |
+
{"current_steps": 4000, "total_steps": 16011, "loss": 3.9257, "lr": 8.704314379665195e-05, "epoch": 0.12033513334636964, "percentage": 24.98, "elapsed_time": "2:16:56", "remaining_time": "6:51:10"}
|
| 44 |
+
{"current_steps": 4096, "total_steps": 16011, "eval_loss": 3.9077117443084717, "epoch": 0.12322317654668251, "percentage": 25.58, "elapsed_time": "2:20:29", "remaining_time": "6:48:41"}
|
| 45 |
+
{"current_steps": 4100, "total_steps": 16011, "loss": 3.9101, "lr": 8.636333809193176e-05, "epoch": 0.12334351168002888, "percentage": 25.61, "elapsed_time": "2:20:41", "remaining_time": "6:48:44"}
|
| 46 |
+
{"current_steps": 4200, "total_steps": 16011, "loss": 3.9048, "lr": 8.566895421445784e-05, "epoch": 0.12635189001368813, "percentage": 26.23, "elapsed_time": "2:24:05", "remaining_time": "6:45:13"}
|
| 47 |
+
{"current_steps": 4300, "total_steps": 16011, "loss": 3.892, "lr": 8.496027054483702e-05, "epoch": 0.12936026834734737, "percentage": 26.86, "elapsed_time": "2:27:30", "remaining_time": "6:41:43"}
|
| 48 |
+
{"current_steps": 4400, "total_steps": 16011, "loss": 3.8769, "lr": 8.423757119650619e-05, "epoch": 0.1323686466810066, "percentage": 27.48, "elapsed_time": "2:30:54", "remaining_time": "6:38:14"}
|
| 49 |
+
{"current_steps": 4500, "total_steps": 16011, "loss": 3.8604, "lr": 8.35011459018305e-05, "epoch": 0.13537702501466584, "percentage": 28.11, "elapsed_time": "2:34:19", "remaining_time": "6:34:45"}
|
| 50 |
+
{"current_steps": 4600, "total_steps": 16011, "loss": 3.8603, "lr": 8.275128989594867e-05, "epoch": 0.13838540334832508, "percentage": 28.73, "elapsed_time": "2:37:43", "remaining_time": "6:31:16"}
|
| 51 |
+
{"current_steps": 4700, "total_steps": 16011, "loss": 3.8484, "lr": 8.198830379841233e-05, "epoch": 0.14139378168198433, "percentage": 29.35, "elapsed_time": "2:41:08", "remaining_time": "6:27:47"}
|
| 52 |
+
{"current_steps": 4800, "total_steps": 16011, "loss": 3.8355, "lr": 8.121249349266684e-05, "epoch": 0.14440216001564357, "percentage": 29.98, "elapsed_time": "2:44:32", "remaining_time": "6:24:19"}
|
| 53 |
+
{"current_steps": 4900, "total_steps": 16011, "loss": 3.8253, "lr": 8.042417000342151e-05, "epoch": 0.14741053834930282, "percentage": 30.6, "elapsed_time": "2:47:57", "remaining_time": "6:20:50"}
|
| 54 |
+
{"current_steps": 5000, "total_steps": 16011, "loss": 3.8184, "lr": 7.962364937195902e-05, "epoch": 0.15041891668296206, "percentage": 31.23, "elapsed_time": "2:51:21", "remaining_time": "6:17:21"}
|
| 55 |
+
{"current_steps": 5100, "total_steps": 16011, "loss": 3.8074, "lr": 7.88112525294334e-05, "epoch": 0.15342729501662128, "percentage": 31.85, "elapsed_time": "2:54:45", "remaining_time": "6:13:52"}
|
| 56 |
+
{"current_steps": 5120, "total_steps": 16011, "eval_loss": 3.8044040203094482, "epoch": 0.15402897068335314, "percentage": 31.98, "elapsed_time": "2:55:43", "remaining_time": "6:13:47"}
|
| 57 |
+
{"current_steps": 5200, "total_steps": 16011, "loss": 3.8047, "lr": 7.798730516820779e-05, "epoch": 0.15643567335028052, "percentage": 32.48, "elapsed_time": "2:58:26", "remaining_time": "6:10:58"}
|
| 58 |
+
{"current_steps": 5300, "total_steps": 16011, "loss": 3.7919, "lr": 7.715213761128337e-05, "epoch": 0.15944405168393977, "percentage": 33.1, "elapsed_time": "3:01:49", "remaining_time": "6:07:27"}
|
| 59 |
+
{"current_steps": 5400, "total_steps": 16011, "loss": 3.7852, "lr": 7.63060846798719e-05, "epoch": 0.16245243001759901, "percentage": 33.73, "elapsed_time": "3:05:13", "remaining_time": "6:03:58"}
|
| 60 |
+
{"current_steps": 5500, "total_steps": 16011, "loss": 3.781, "lr": 7.544948555916485e-05, "epoch": 0.16546080835125826, "percentage": 34.35, "elapsed_time": "3:08:37", "remaining_time": "6:00:29"}
|
| 61 |
+
{"current_steps": 5600, "total_steps": 16011, "loss": 3.7734, "lr": 7.4582683662353e-05, "epoch": 0.1684691866849175, "percentage": 34.98, "elapsed_time": "3:12:01", "remaining_time": "5:56:59"}
|
| 62 |
+
{"current_steps": 5700, "total_steps": 16011, "loss": 3.7631, "lr": 7.370602649295112e-05, "epoch": 0.17147756501857675, "percentage": 35.6, "elapsed_time": "3:15:25", "remaining_time": "5:53:31"}
|
| 63 |
+
{"current_steps": 5800, "total_steps": 16011, "loss": 3.7529, "lr": 7.281986550548268e-05, "epoch": 0.17448594335223597, "percentage": 36.23, "elapsed_time": "3:18:49", "remaining_time": "5:50:02"}
|
| 64 |
+
{"current_steps": 5900, "total_steps": 16011, "loss": 3.742, "lr": 7.192455596458073e-05, "epoch": 0.1774943216858952, "percentage": 36.85, "elapsed_time": "3:22:14", "remaining_time": "5:46:34"}
|
| 65 |
+
{"current_steps": 6000, "total_steps": 16011, "loss": 3.7437, "lr": 7.102045680256113e-05, "epoch": 0.18050270001955446, "percentage": 37.47, "elapsed_time": "3:25:38", "remaining_time": "5:43:06"}
|
| 66 |
+
{"current_steps": 6100, "total_steps": 16011, "loss": 3.7379, "lr": 7.010793047552565e-05, "epoch": 0.1835110783532137, "percentage": 38.1, "elapsed_time": "3:29:03", "remaining_time": "5:39:39"}
|
| 67 |
+
{"current_steps": 6144, "total_steps": 16011, "eval_loss": 3.7293174266815186, "epoch": 0.18483476482002376, "percentage": 38.37, "elapsed_time": "3:30:50", "remaining_time": "5:38:35"}
|
| 68 |
+
{"current_steps": 6200, "total_steps": 16011, "loss": 3.73, "lr": 6.918734281805204e-05, "epoch": 0.18651945668687295, "percentage": 38.72, "elapsed_time": "3:32:48", "remaining_time": "5:36:44"}
|
| 69 |
+
{"current_steps": 6300, "total_steps": 16011, "loss": 3.7287, "lr": 6.825906289652998e-05, "epoch": 0.1895278350205322, "percentage": 39.35, "elapsed_time": "3:36:12", "remaining_time": "5:33:16"}
|
| 70 |
+
{"current_steps": 6400, "total_steps": 16011, "loss": 3.7126, "lr": 6.732346286120116e-05, "epoch": 0.1925362133541914, "percentage": 39.97, "elapsed_time": "3:39:36", "remaining_time": "5:29:48"}
|
| 71 |
+
{"current_steps": 6500, "total_steps": 16011, "loss": 3.7079, "lr": 6.638091779696314e-05, "epoch": 0.19554459168785066, "percentage": 40.6, "elapsed_time": "3:43:01", "remaining_time": "5:26:20"}
|
| 72 |
+
{"current_steps": 6600, "total_steps": 16011, "loss": 3.703, "lr": 6.543180557299673e-05, "epoch": 0.1985529700215099, "percentage": 41.22, "elapsed_time": "3:46:25", "remaining_time": "5:22:52"}
|
| 73 |
+
{"current_steps": 6700, "total_steps": 16011, "loss": 3.7042, "lr": 6.447650669127692e-05, "epoch": 0.20156134835516915, "percentage": 41.85, "elapsed_time": "3:49:50", "remaining_time": "5:19:24"}
|
| 74 |
+
{"current_steps": 6800, "total_steps": 16011, "loss": 3.7031, "lr": 6.35154041340287e-05, "epoch": 0.2045697266888284, "percentage": 42.47, "elapsed_time": "3:53:14", "remaining_time": "5:15:55"}
|
| 75 |
+
{"current_steps": 6900, "total_steps": 16011, "loss": 3.6848, "lr": 6.254888321018817e-05, "epoch": 0.20757810502248764, "percentage": 43.1, "elapsed_time": "3:56:38", "remaining_time": "5:12:27"}
|
| 76 |
+
{"current_steps": 7000, "total_steps": 16011, "loss": 3.6811, "lr": 6.157733140093107e-05, "epoch": 0.21058648335614688, "percentage": 43.72, "elapsed_time": "4:00:02", "remaining_time": "5:08:59"}
|
| 77 |
+
{"current_steps": 7100, "total_steps": 16011, "loss": 3.6719, "lr": 6.060113820433041e-05, "epoch": 0.2135948616898061, "percentage": 44.34, "elapsed_time": "4:03:25", "remaining_time": "5:05:30"}
|
| 78 |
+
{"current_steps": 7168, "total_steps": 16011, "eval_loss": 3.674034595489502, "epoch": 0.2156405589566944, "percentage": 44.77, "elapsed_time": "4:06:00", "remaining_time": "5:03:30"}
|
| 79 |
+
{"current_steps": 7200, "total_steps": 16011, "loss": 3.6764, "lr": 5.962069497920546e-05, "epoch": 0.21660324002346534, "percentage": 44.97, "elapsed_time": "4:07:05", "remaining_time": "5:02:23"}
|
| 80 |
+
{"current_steps": 7300, "total_steps": 16011, "loss": 3.6707, "lr": 5.863639478822488e-05, "epoch": 0.2196116183571246, "percentage": 45.59, "elapsed_time": "4:10:29", "remaining_time": "4:58:54"}
|
| 81 |
+
{"current_steps": 7400, "total_steps": 16011, "loss": 3.6606, "lr": 5.764863224032663e-05, "epoch": 0.22261999669078383, "percentage": 46.22, "elapsed_time": "4:13:53", "remaining_time": "4:55:26"}
|
| 82 |
+
{"current_steps": 7500, "total_steps": 16011, "loss": 3.6531, "lr": 5.66578033325181e-05, "epoch": 0.22562837502444308, "percentage": 46.84, "elapsed_time": "4:17:18", "remaining_time": "4:51:59"}
|
| 83 |
+
{"current_steps": 7600, "total_steps": 16011, "loss": 3.6599, "lr": 5.566430529111969e-05, "epoch": 0.22863675335810232, "percentage": 47.47, "elapsed_time": "4:20:42", "remaining_time": "4:48:32"}
|
| 84 |
+
{"current_steps": 7700, "total_steps": 16011, "loss": 3.6551, "lr": 5.4668536412515534e-05, "epoch": 0.23164513169176157, "percentage": 48.09, "elapsed_time": "4:24:07", "remaining_time": "4:45:04"}
|
| 85 |
+
{"current_steps": 7800, "total_steps": 16011, "loss": 3.6517, "lr": 5.367089590347535e-05, "epoch": 0.2346535100254208, "percentage": 48.72, "elapsed_time": "4:27:30", "remaining_time": "4:41:36"}
|
| 86 |
+
{"current_steps": 7900, "total_steps": 16011, "loss": 3.639, "lr": 5.2671783721111145e-05, "epoch": 0.23766188835908003, "percentage": 49.34, "elapsed_time": "4:30:54", "remaining_time": "4:38:08"}
|
| 87 |
+
{"current_steps": 8000, "total_steps": 16011, "loss": 3.6392, "lr": 5.1671600412533284e-05, "epoch": 0.24067026669273928, "percentage": 49.97, "elapsed_time": "4:34:18", "remaining_time": "4:34:40"}
|
| 88 |
+
{"current_steps": 8100, "total_steps": 16011, "loss": 3.6267, "lr": 5.067074695426984e-05, "epoch": 0.24367864502639852, "percentage": 50.59, "elapsed_time": "4:37:42", "remaining_time": "4:31:14"}
|
| 89 |
+
{"current_steps": 8192, "total_steps": 16011, "eval_loss": 3.629561185836792, "epoch": 0.24644635309336502, "percentage": 51.16, "elapsed_time": "4:41:07", "remaining_time": "4:28:19"}
|
| 90 |
+
{"current_steps": 8200, "total_steps": 16011, "loss": 3.6274, "lr": 4.966962459151393e-05, "epoch": 0.24668702336005777, "percentage": 51.21, "elapsed_time": "4:41:27", "remaining_time": "4:28:06"}
|
| 91 |
+
{"current_steps": 8300, "total_steps": 16011, "loss": 3.6255, "lr": 4.866863467726335e-05, "epoch": 0.249695401693717, "percentage": 51.84, "elapsed_time": "4:44:52", "remaining_time": "4:24:39"}
|
| 92 |
+
{"current_steps": 8400, "total_steps": 16011, "loss": 3.622, "lr": 4.7668178511416815e-05, "epoch": 0.25270378002737626, "percentage": 52.46, "elapsed_time": "4:48:16", "remaining_time": "4:21:11"}
|
| 93 |
+
{"current_steps": 8500, "total_steps": 16011, "loss": 3.6162, "lr": 4.666865717989178e-05, "epoch": 0.2557121583610355, "percentage": 53.09, "elapsed_time": "4:51:40", "remaining_time": "4:17:44"}
|
| 94 |
+
{"current_steps": 8600, "total_steps": 16011, "loss": 3.6235, "lr": 4.567047139382766e-05, "epoch": 0.25872053669469475, "percentage": 53.71, "elapsed_time": "4:55:05", "remaining_time": "4:14:17"}
|
| 95 |
+
{"current_steps": 8700, "total_steps": 16011, "loss": 3.6107, "lr": 4.4674021328939746e-05, "epoch": 0.261728915028354, "percentage": 54.34, "elapsed_time": "4:58:30", "remaining_time": "4:10:50"}
|
| 96 |
+
{"current_steps": 8800, "total_steps": 16011, "loss": 3.611, "lr": 4.3679706465087344e-05, "epoch": 0.2647372933620132, "percentage": 54.96, "elapsed_time": "5:01:55", "remaining_time": "4:07:24"}
|
| 97 |
+
{"current_steps": 8900, "total_steps": 16011, "loss": 3.6025, "lr": 4.268792542612125e-05, "epoch": 0.2677456716956724, "percentage": 55.59, "elapsed_time": "5:05:19", "remaining_time": "4:03:56"}
|
| 98 |
+
{"current_steps": 9000, "total_steps": 16011, "loss": 3.6052, "lr": 4.169907582007412e-05, "epoch": 0.27075405002933167, "percentage": 56.21, "elapsed_time": "5:08:43", "remaining_time": "4:00:30"}
|
| 99 |
+
{"current_steps": 9100, "total_steps": 16011, "loss": 3.5891, "lr": 4.071355407975825e-05, "epoch": 0.2737624283629909, "percentage": 56.84, "elapsed_time": "5:12:07", "remaining_time": "3:57:02"}
|
| 100 |
+
{"current_steps": 9200, "total_steps": 16011, "loss": 3.5901, "lr": 3.9731755303834453e-05, "epoch": 0.27677080669665016, "percentage": 57.46, "elapsed_time": "5:15:31", "remaining_time": "3:53:35"}
|
| 101 |
+
{"current_steps": 9216, "total_steps": 16011, "eval_loss": 3.5964441299438477, "epoch": 0.27725214723003566, "percentage": 57.56, "elapsed_time": "5:16:21", "remaining_time": "3:53:15"}
|
| 102 |
+
{"current_steps": 9300, "total_steps": 16011, "loss": 3.5833, "lr": 3.875407309841582e-05, "epoch": 0.2797791850303094, "percentage": 58.09, "elapsed_time": "5:19:12", "remaining_time": "3:50:20"}
|
| 103 |
+
{"current_steps": 9400, "total_steps": 16011, "loss": 3.5885, "lr": 3.778089941926976e-05, "epoch": 0.28278756336396865, "percentage": 58.71, "elapsed_time": "5:22:35", "remaining_time": "3:46:52"}
|
| 104 |
+
{"current_steps": 9500, "total_steps": 16011, "loss": 3.5791, "lr": 3.681262441468189e-05, "epoch": 0.2857959416976279, "percentage": 59.33, "elapsed_time": "5:25:59", "remaining_time": "3:43:25"}
|
| 105 |
+
{"current_steps": 9600, "total_steps": 16011, "loss": 3.5907, "lr": 3.584963626904423e-05, "epoch": 0.28880432003128714, "percentage": 59.96, "elapsed_time": "5:29:23", "remaining_time": "3:39:58"}
|
| 106 |
+
{"current_steps": 9700, "total_steps": 16011, "loss": 3.5757, "lr": 3.4892321047231e-05, "epoch": 0.2918126983649464, "percentage": 60.58, "elapsed_time": "5:32:47", "remaining_time": "3:36:31"}
|
| 107 |
+
{"current_steps": 9800, "total_steps": 16011, "loss": 3.5781, "lr": 3.3941062539824027e-05, "epoch": 0.29482107669860563, "percentage": 61.21, "elapsed_time": "5:36:11", "remaining_time": "3:33:04"}
|
| 108 |
+
{"current_steps": 9900, "total_steps": 16011, "loss": 3.5735, "lr": 3.299624210924994e-05, "epoch": 0.2978294550322649, "percentage": 61.83, "elapsed_time": "5:39:35", "remaining_time": "3:29:37"}
|
| 109 |
+
{"current_steps": 10000, "total_steps": 16011, "loss": 3.5721, "lr": 3.205823853689076e-05, "epoch": 0.3008378333659241, "percentage": 62.46, "elapsed_time": "5:42:59", "remaining_time": "3:26:10"}
|
| 110 |
+
{"current_steps": 10100, "total_steps": 16011, "loss": 3.5775, "lr": 3.112742787122936e-05, "epoch": 0.3038462116995833, "percentage": 63.08, "elapsed_time": "5:46:23", "remaining_time": "3:22:43"}
|
| 111 |
+
{"current_steps": 10200, "total_steps": 16011, "loss": 3.5696, "lr": 3.020418327709045e-05, "epoch": 0.30685459003324256, "percentage": 63.71, "elapsed_time": "5:49:47", "remaining_time": "3:19:16"}
|
| 112 |
+
{"current_steps": 10240, "total_steps": 16011, "eval_loss": 3.5699303150177, "epoch": 0.3080579413667063, "percentage": 63.96, "elapsed_time": "5:51:26", "remaining_time": "3:18:03"}
|
| 113 |
+
{"current_steps": 10300, "total_steps": 16011, "loss": 3.5676, "lr": 2.928887488603769e-05, "epoch": 0.3098629683669018, "percentage": 64.33, "elapsed_time": "5:53:32", "remaining_time": "3:16:01"}
|
| 114 |
+
{"current_steps": 10400, "total_steps": 16011, "loss": 3.5622, "lr": 2.8381869647986746e-05, "epoch": 0.31287134670056105, "percentage": 64.96, "elapsed_time": "5:56:56", "remaining_time": "3:12:34"}
|
| 115 |
+
{"current_steps": 10500, "total_steps": 16011, "loss": 3.5639, "lr": 2.7483531184094124e-05, "epoch": 0.3158797250342203, "percentage": 65.58, "elapsed_time": "6:00:20", "remaining_time": "3:09:07"}
|
| 116 |
+
{"current_steps": 10600, "total_steps": 16011, "loss": 3.562, "lr": 2.659421964098008e-05, "epoch": 0.31888810336787954, "percentage": 66.2, "elapsed_time": "6:03:44", "remaining_time": "3:05:40"}
|
| 117 |
+
{"current_steps": 10700, "total_steps": 16011, "loss": 3.5602, "lr": 2.5714291546345044e-05, "epoch": 0.3218964817015388, "percentage": 66.83, "elapsed_time": "6:07:08", "remaining_time": "3:02:14"}
|
| 118 |
+
{"current_steps": 10800, "total_steps": 16011, "loss": 3.5584, "lr": 2.484409966603632e-05, "epoch": 0.32490486003519803, "percentage": 67.45, "elapsed_time": "6:10:33", "remaining_time": "2:58:47"}
|
| 119 |
+
{"current_steps": 10900, "total_steps": 16011, "loss": 3.5482, "lr": 2.3983992862623362e-05, "epoch": 0.3279132383688573, "percentage": 68.08, "elapsed_time": "6:13:57", "remaining_time": "2:55:20"}
|
| 120 |
+
{"current_steps": 11000, "total_steps": 16011, "loss": 3.5534, "lr": 2.313431595553762e-05, "epoch": 0.3309216167025165, "percentage": 68.7, "elapsed_time": "6:17:21", "remaining_time": "2:51:54"}
|
| 121 |
+
{"current_steps": 11100, "total_steps": 16011, "loss": 3.551, "lr": 2.2295409582833425e-05, "epoch": 0.33392999503617576, "percentage": 69.33, "elapsed_time": "6:20:45", "remaining_time": "2:48:27"}
|
| 122 |
+
{"current_steps": 11200, "total_steps": 16011, "loss": 3.5561, "lr": 2.1467610064625133e-05, "epoch": 0.336938373369835, "percentage": 69.95, "elapsed_time": "6:24:10", "remaining_time": "2:45:01"}
|
| 123 |
+
{"current_steps": 11264, "total_steps": 16011, "eval_loss": 3.550001859664917, "epoch": 0.3388637355033769, "percentage": 70.35, "elapsed_time": "6:26:38", "remaining_time": "2:42:56"}
|
| 124 |
+
{"current_steps": 11300, "total_steps": 16011, "loss": 3.5471, "lr": 2.0651249268255396e-05, "epoch": 0.33994675170349425, "percentage": 70.58, "elapsed_time": "6:27:51", "remaining_time": "2:41:41"}
|
| 125 |
+
{"current_steps": 11400, "total_steps": 16011, "loss": 3.547, "lr": 1.9846654475248443e-05, "epoch": 0.3429551300371535, "percentage": 71.2, "elapsed_time": "6:31:15", "remaining_time": "2:38:15"}
|
| 126 |
+
{"current_steps": 11500, "total_steps": 16011, "loss": 3.5484, "lr": 1.9054148250102126e-05, "epoch": 0.3459635083708127, "percentage": 71.83, "elapsed_time": "6:34:39", "remaining_time": "2:34:48"}
|
| 127 |
+
{"current_steps": 11600, "total_steps": 16011, "loss": 3.5431, "lr": 1.8274048310970558e-05, "epoch": 0.34897188670447193, "percentage": 72.45, "elapsed_time": "6:38:03", "remaining_time": "2:31:21"}
|
| 128 |
+
{"current_steps": 11700, "total_steps": 16011, "loss": 3.5303, "lr": 1.7506667402290234e-05, "epoch": 0.3519802650381312, "percentage": 73.07, "elapsed_time": "6:41:27", "remaining_time": "2:27:55"}
|
| 129 |
+
{"current_steps": 11800, "total_steps": 16011, "loss": 3.5363, "lr": 1.6752313169399547e-05, "epoch": 0.3549886433717904, "percentage": 73.7, "elapsed_time": "6:44:51", "remaining_time": "2:24:28"}
|
| 130 |
+
{"current_steps": 11900, "total_steps": 16011, "loss": 3.5364, "lr": 1.601128803520306e-05, "epoch": 0.35799702170544967, "percentage": 74.32, "elapsed_time": "6:48:15", "remaining_time": "2:21:02"}
|
| 131 |
+
{"current_steps": 12000, "total_steps": 16011, "loss": 3.5384, "lr": 1.5283889078929137e-05, "epoch": 0.3610054000391089, "percentage": 74.95, "elapsed_time": "6:51:38", "remaining_time": "2:17:35"}
|
| 132 |
+
{"current_steps": 12100, "total_steps": 16011, "loss": 3.537, "lr": 1.4570407917030005e-05, "epoch": 0.36401377837276816, "percentage": 75.57, "elapsed_time": "6:55:03", "remaining_time": "2:14:09"}
|
| 133 |
+
{"current_steps": 12200, "total_steps": 16011, "loss": 3.5263, "lr": 1.387113058627183e-05, "epoch": 0.3670221567064274, "percentage": 76.2, "elapsed_time": "6:58:27", "remaining_time": "2:10:42"}
|
| 134 |
+
{"current_steps": 12288, "total_steps": 16011, "eval_loss": 3.5356359481811523, "epoch": 0.3696695296400475, "percentage": 76.75, "elapsed_time": "7:01:44", "remaining_time": "2:07:46"}
|
| 135 |
+
{"current_steps": 12300, "total_steps": 16011, "loss": 3.5311, "lr": 1.3186337429061708e-05, "epoch": 0.37003053504008665, "percentage": 76.82, "elapsed_time": "7:02:12", "remaining_time": "2:07:23"}
|
| 136 |
+
{"current_steps": 12400, "total_steps": 16011, "loss": 3.5335, "lr": 1.2516302981057498e-05, "epoch": 0.3730389133737459, "percentage": 77.45, "elapsed_time": "7:05:37", "remaining_time": "2:03:56"}
|
| 137 |
+
{"current_steps": 12500, "total_steps": 16011, "loss": 3.5342, "lr": 1.186129586110578e-05, "epoch": 0.37604729170740514, "percentage": 78.07, "elapsed_time": "7:09:01", "remaining_time": "2:00:30"}
|
| 138 |
+
{"current_steps": 12600, "total_steps": 16011, "loss": 3.531, "lr": 1.1221578663551536e-05, "epoch": 0.3790556700410644, "percentage": 78.7, "elapsed_time": "7:12:26", "remaining_time": "1:57:04"}
|
| 139 |
+
{"current_steps": 12700, "total_steps": 16011, "loss": 3.525, "lr": 1.0597407852963476e-05, "epoch": 0.38206404837472363, "percentage": 79.32, "elapsed_time": "7:15:51", "remaining_time": "1:53:37"}
|
| 140 |
+
{"current_steps": 12800, "total_steps": 16011, "loss": 3.5278, "lr": 9.98903366131651e-06, "epoch": 0.3850724267083828, "percentage": 79.95, "elapsed_time": "7:19:16", "remaining_time": "1:50:11"}
|
| 141 |
+
{"current_steps": 12900, "total_steps": 16011, "loss": 3.5294, "lr": 9.39669998767294e-06, "epoch": 0.38808080504204207, "percentage": 80.57, "elapsed_time": "7:22:40", "remaining_time": "1:46:45"}
|
| 142 |
+
{"current_steps": 13000, "total_steps": 16011, "loss": 3.5207, "lr": 8.820644300402764e-06, "epoch": 0.3910891833757013, "percentage": 81.19, "elapsed_time": "7:26:04", "remaining_time": "1:43:19"}
|
| 143 |
+
{"current_steps": 13100, "total_steps": 16011, "loss": 3.524, "lr": 8.261097541981678e-06, "epoch": 0.39409756170936056, "percentage": 81.82, "elapsed_time": "7:29:28", "remaining_time": "1:39:52"}
|
| 144 |
+
{"current_steps": 13200, "total_steps": 16011, "loss": 3.5204, "lr": 7.718284036405666e-06, "epoch": 0.3971059400430198, "percentage": 82.44, "elapsed_time": "7:32:52", "remaining_time": "1:36:26"}
|
| 145 |
+
{"current_steps": 13300, "total_steps": 16011, "loss": 3.5247, "lr": 7.192421399258803e-06, "epoch": 0.40011431837667905, "percentage": 83.07, "elapsed_time": "7:36:16", "remaining_time": "1:33:00"}
|
| 146 |
+
{"current_steps": 13312, "total_steps": 16011, "eval_loss": 3.5258023738861084, "epoch": 0.4004753237767182, "percentage": 83.14, "elapsed_time": "7:36:58", "remaining_time": "1:32:39"}
|
| 147 |
+
{"current_steps": 13400, "total_steps": 16011, "loss": 3.5247, "lr": 6.683720450470471e-06, "epoch": 0.4031226967103383, "percentage": 83.69, "elapsed_time": "7:39:58", "remaining_time": "1:29:37"}
|
| 148 |
+
{"current_steps": 13500, "total_steps": 16011, "loss": 3.5208, "lr": 6.192385129797057e-06, "epoch": 0.40613107504399754, "percentage": 84.32, "elapsed_time": "7:43:22", "remaining_time": "1:26:11"}
|
| 149 |
+
{"current_steps": 13600, "total_steps": 16011, "loss": 3.5168, "lr": 5.7186124150618355e-06, "epoch": 0.4091394533776568, "percentage": 84.94, "elapsed_time": "7:46:46", "remaining_time": "1:22:45"}
|
| 150 |
+
{"current_steps": 13700, "total_steps": 16011, "loss": 3.5276, "lr": 5.262592243185998e-06, "epoch": 0.412147831711316, "percentage": 85.57, "elapsed_time": "7:50:10", "remaining_time": "1:19:18"}
|
| 151 |
+
{"current_steps": 13800, "total_steps": 16011, "loss": 3.5207, "lr": 4.824507434042369e-06, "epoch": 0.41515621004497527, "percentage": 86.19, "elapsed_time": "7:53:35", "remaining_time": "1:15:52"}
|
| 152 |
+
{"current_steps": 13900, "total_steps": 16011, "loss": 3.522, "lr": 4.4045336171622985e-06, "epoch": 0.4181645883786345, "percentage": 86.82, "elapsed_time": "7:56:59", "remaining_time": "1:12:26"}
|
| 153 |
+
{"current_steps": 14000, "total_steps": 16011, "loss": 3.5178, "lr": 4.002839161325356e-06, "epoch": 0.42117296671229376, "percentage": 87.44, "elapsed_time": "8:00:23", "remaining_time": "1:09:00"}
|
| 154 |
+
{"current_steps": 14100, "total_steps": 16011, "loss": 3.5185, "lr": 3.6195851070596586e-06, "epoch": 0.424181345045953, "percentage": 88.06, "elapsed_time": "8:03:47", "remaining_time": "1:05:34"}
|
| 155 |
+
{"current_steps": 14200, "total_steps": 16011, "loss": 3.5182, "lr": 3.2549251020803307e-06, "epoch": 0.4271897233796122, "percentage": 88.69, "elapsed_time": "8:07:11", "remaining_time": "1:02:07"}
|
| 156 |
+
{"current_steps": 14300, "total_steps": 16011, "loss": 3.521, "lr": 2.9090053396916172e-06, "epoch": 0.43019810171327144, "percentage": 89.31, "elapsed_time": "8:10:35", "remaining_time": "0:58:41"}
|
| 157 |
+
{"current_steps": 14336, "total_steps": 16011, "eval_loss": 3.520242929458618, "epoch": 0.4312811179133888, "percentage": 89.54, "elapsed_time": "8:12:06", "remaining_time": "0:57:29"}
|
| 158 |
+
{"current_steps": 14400, "total_steps": 16011, "loss": 3.5214, "lr": 2.5819645001775474e-06, "epoch": 0.4332064800469307, "percentage": 89.94, "elapsed_time": "8:14:20", "remaining_time": "0:55:18"}
|
| 159 |
+
{"current_steps": 14500, "total_steps": 16011, "loss": 3.5155, "lr": 2.2739336952046e-06, "epoch": 0.43621485838058993, "percentage": 90.56, "elapsed_time": "8:17:44", "remaining_time": "0:51:52"}
|
| 160 |
+
{"current_steps": 14600, "total_steps": 16011, "loss": 3.5169, "lr": 1.9850364152586344e-06, "epoch": 0.4392232367142492, "percentage": 91.19, "elapsed_time": "8:21:08", "remaining_time": "0:48:25"}
|
| 161 |
+
{"current_steps": 14700, "total_steps": 16011, "loss": 3.5171, "lr": 1.7153884801371756e-06, "epoch": 0.4422316150479084, "percentage": 91.81, "elapsed_time": "8:24:33", "remaining_time": "0:44:59"}
|
| 162 |
+
{"current_steps": 14800, "total_steps": 16011, "loss": 3.5133, "lr": 1.4650979925169117e-06, "epoch": 0.44523999338156767, "percentage": 92.44, "elapsed_time": "8:27:57", "remaining_time": "0:41:33"}
|
| 163 |
+
{"current_steps": 14900, "total_steps": 16011, "loss": 3.5186, "lr": 1.234265294614978e-06, "epoch": 0.4482483717152269, "percentage": 93.06, "elapsed_time": "8:31:21", "remaining_time": "0:38:07"}
|
| 164 |
+
{"current_steps": 15000, "total_steps": 16011, "loss": 3.522, "lr": 1.0229829279615133e-06, "epoch": 0.45125675004888616, "percentage": 93.69, "elapsed_time": "8:34:45", "remaining_time": "0:34:41"}
|
| 165 |
+
{"current_steps": 15100, "total_steps": 16011, "loss": 3.5216, "lr": 8.313355962994041e-07, "epoch": 0.4542651283825454, "percentage": 94.31, "elapsed_time": "8:38:09", "remaining_time": "0:31:15"}
|
| 166 |
+
{"current_steps": 15200, "total_steps": 16011, "loss": 3.5198, "lr": 6.594001316263587e-07, "epoch": 0.45727350671620465, "percentage": 94.93, "elapsed_time": "8:41:33", "remaining_time": "0:27:49"}
|
| 167 |
+
{"current_steps": 15300, "total_steps": 16011, "loss": 3.5164, "lr": 5.072454633927026e-07, "epoch": 0.4602818850498639, "percentage": 95.56, "elapsed_time": "8:44:57", "remaining_time": "0:24:23"}
|
| 168 |
+
{"current_steps": 15360, "total_steps": 16011, "eval_loss": 3.518115520477295, "epoch": 0.4620869120500594, "percentage": 95.93, "elapsed_time": "8:47:18", "remaining_time": "0:22:20"}
|
| 169 |
+
{"current_steps": 15400, "total_steps": 16011, "loss": 3.514, "lr": 3.7493259086733845e-07, "epoch": 0.46329026338352314, "percentage": 96.18, "elapsed_time": "8:48:39", "remaining_time": "0:20:58"}
|
| 170 |
+
{"current_steps": 15500, "total_steps": 16011, "loss": 3.5173, "lr": 2.625145586829858e-07, "epoch": 0.4662986417171823, "percentage": 96.81, "elapsed_time": "8:52:03", "remaining_time": "0:17:32"}
|
| 171 |
+
{"current_steps": 15600, "total_steps": 16011, "loss": 3.5157, "lr": 1.7003643557039762e-07, "epoch": 0.4693070200508416, "percentage": 97.43, "elapsed_time": "8:55:27", "remaining_time": "0:14:06"}
|
| 172 |
+
{"current_steps": 15700, "total_steps": 16011, "loss": 3.5182, "lr": 9.753529629017988e-08, "epoch": 0.4723153983845008, "percentage": 98.06, "elapsed_time": "8:58:51", "remaining_time": "0:10:40"}
|
| 173 |
+
{"current_steps": 15800, "total_steps": 16011, "loss": 3.5106, "lr": 4.504020676941445e-08, "epoch": 0.47532377671816006, "percentage": 98.68, "elapsed_time": "9:02:16", "remaining_time": "0:07:14"}
|
| 174 |
+
{"current_steps": 15900, "total_steps": 16011, "loss": 3.5156, "lr": 1.2572212449019072e-08, "epoch": 0.4783321550518193, "percentage": 99.31, "elapsed_time": "9:05:40", "remaining_time": "0:03:48"}
|
| 175 |
+
{"current_steps": 16000, "total_steps": 16011, "loss": 3.5122, "lr": 1.4432984660195382e-10, "epoch": 0.48134053338547855, "percentage": 99.93, "elapsed_time": "9:09:04", "remaining_time": "0:00:22"}
|
| 176 |
+
{"current_steps": 16011, "total_steps": 16011, "epoch": 0.48167145500218106, "percentage": 100.0, "elapsed_time": "9:09:31", "remaining_time": "0:00:00"}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/trainer_state.json
ADDED
|
@@ -0,0 +1,1283 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.48167145500218106,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 16011,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0030083783336592408,
|
| 14 |
+
"grad_norm": 1.7087676525115967,
|
| 15 |
+
"learning_rate": 3.08411214953271e-05,
|
| 16 |
+
"loss": 10.6918,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.0060167566673184816,
|
| 21 |
+
"grad_norm": 1.1703847646713257,
|
| 22 |
+
"learning_rate": 6.199376947040498e-05,
|
| 23 |
+
"loss": 7.9189,
|
| 24 |
+
"step": 200
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.009025135000977723,
|
| 28 |
+
"grad_norm": 0.8664107322692871,
|
| 29 |
+
"learning_rate": 9.314641744548286e-05,
|
| 30 |
+
"loss": 6.7175,
|
| 31 |
+
"step": 300
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.012033513334636963,
|
| 35 |
+
"grad_norm": 1.0361624956130981,
|
| 36 |
+
"learning_rate": 9.999390218499588e-05,
|
| 37 |
+
"loss": 6.1791,
|
| 38 |
+
"step": 400
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.015041891668296205,
|
| 42 |
+
"grad_norm": 1.3140676021575928,
|
| 43 |
+
"learning_rate": 9.996824677076817e-05,
|
| 44 |
+
"loss": 5.8595,
|
| 45 |
+
"step": 500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.018050270001955446,
|
| 49 |
+
"grad_norm": 1.210119605064392,
|
| 50 |
+
"learning_rate": 9.992255893455041e-05,
|
| 51 |
+
"loss": 5.637,
|
| 52 |
+
"step": 600
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.021058648335614688,
|
| 56 |
+
"grad_norm": 0.9253405332565308,
|
| 57 |
+
"learning_rate": 9.9856856992735e-05,
|
| 58 |
+
"loss": 5.4491,
|
| 59 |
+
"step": 700
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.024067026669273926,
|
| 63 |
+
"grad_norm": 0.9484109878540039,
|
| 64 |
+
"learning_rate": 9.977116728543007e-05,
|
| 65 |
+
"loss": 5.3055,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.027075405002933168,
|
| 70 |
+
"grad_norm": 0.867859423160553,
|
| 71 |
+
"learning_rate": 9.966552416589973e-05,
|
| 72 |
+
"loss": 5.1579,
|
| 73 |
+
"step": 900
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.03008378333659241,
|
| 77 |
+
"grad_norm": 0.8659229278564453,
|
| 78 |
+
"learning_rate": 9.953996998679162e-05,
|
| 79 |
+
"loss": 5.0361,
|
| 80 |
+
"step": 1000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.030805794136670627,
|
| 84 |
+
"eval_loss": 4.949322700500488,
|
| 85 |
+
"eval_runtime": 17.3254,
|
| 86 |
+
"eval_samples_per_second": 244.613,
|
| 87 |
+
"eval_steps_per_second": 15.295,
|
| 88 |
+
"step": 1024
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.03309216167025165,
|
| 92 |
+
"grad_norm": 0.8792940974235535,
|
| 93 |
+
"learning_rate": 9.939455508315775e-05,
|
| 94 |
+
"loss": 4.9213,
|
| 95 |
+
"step": 1100
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 0.03610054000391089,
|
| 99 |
+
"grad_norm": 0.8531689047813416,
|
| 100 |
+
"learning_rate": 9.92293377522749e-05,
|
| 101 |
+
"loss": 4.8299,
|
| 102 |
+
"step": 1200
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.03910891833757013,
|
| 106 |
+
"grad_norm": 0.845849335193634,
|
| 107 |
+
"learning_rate": 9.904438423027316e-05,
|
| 108 |
+
"loss": 4.7423,
|
| 109 |
+
"step": 1300
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 0.042117296671229376,
|
| 113 |
+
"grad_norm": 0.6916408538818359,
|
| 114 |
+
"learning_rate": 9.883976866558158e-05,
|
| 115 |
+
"loss": 4.662,
|
| 116 |
+
"step": 1400
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.045125675004888614,
|
| 120 |
+
"grad_norm": 0.6636393666267395,
|
| 121 |
+
"learning_rate": 9.86155730892019e-05,
|
| 122 |
+
"loss": 4.5986,
|
| 123 |
+
"step": 1500
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.04813405333854785,
|
| 127 |
+
"grad_norm": 0.8140773773193359,
|
| 128 |
+
"learning_rate": 9.837188738182205e-05,
|
| 129 |
+
"loss": 4.5433,
|
| 130 |
+
"step": 1600
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.0511424316722071,
|
| 134 |
+
"grad_norm": 0.7255787253379822,
|
| 135 |
+
"learning_rate": 9.81088092377827e-05,
|
| 136 |
+
"loss": 4.4899,
|
| 137 |
+
"step": 1700
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.054150810005866336,
|
| 141 |
+
"grad_norm": 0.7826430201530457,
|
| 142 |
+
"learning_rate": 9.782644412591135e-05,
|
| 143 |
+
"loss": 4.443,
|
| 144 |
+
"step": 1800
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.05715918833952558,
|
| 148 |
+
"grad_norm": 0.7264050245285034,
|
| 149 |
+
"learning_rate": 9.752490524723946e-05,
|
| 150 |
+
"loss": 4.3915,
|
| 151 |
+
"step": 1900
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.06016756667318482,
|
| 155 |
+
"grad_norm": 0.6457399129867554,
|
| 156 |
+
"learning_rate": 9.720431348961991e-05,
|
| 157 |
+
"loss": 4.3616,
|
| 158 |
+
"step": 2000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.061611588273341254,
|
| 162 |
+
"eval_loss": 4.325469493865967,
|
| 163 |
+
"eval_runtime": 17.3482,
|
| 164 |
+
"eval_samples_per_second": 244.29,
|
| 165 |
+
"eval_steps_per_second": 15.275,
|
| 166 |
+
"step": 2048
|
| 167 |
+
},
|
| 168 |
+
{
|
| 169 |
+
"epoch": 0.06317594500684406,
|
| 170 |
+
"grad_norm": 0.6707364320755005,
|
| 171 |
+
"learning_rate": 9.686479737926266e-05,
|
| 172 |
+
"loss": 4.3253,
|
| 173 |
+
"step": 2100
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 0.0661843233405033,
|
| 177 |
+
"grad_norm": 0.701897382736206,
|
| 178 |
+
"learning_rate": 9.650649302920819e-05,
|
| 179 |
+
"loss": 4.2853,
|
| 180 |
+
"step": 2200
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.06919270167416254,
|
| 184 |
+
"grad_norm": 0.6544104814529419,
|
| 185 |
+
"learning_rate": 9.612954408475934e-05,
|
| 186 |
+
"loss": 4.261,
|
| 187 |
+
"step": 2300
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 0.07220108000782179,
|
| 191 |
+
"grad_norm": 0.6391928195953369,
|
| 192 |
+
"learning_rate": 9.573410166589356e-05,
|
| 193 |
+
"loss": 4.229,
|
| 194 |
+
"step": 2400
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.07520945834148103,
|
| 198 |
+
"grad_norm": 0.6567296981811523,
|
| 199 |
+
"learning_rate": 9.532032430667834e-05,
|
| 200 |
+
"loss": 4.2043,
|
| 201 |
+
"step": 2500
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.07821783667514026,
|
| 205 |
+
"grad_norm": 0.6132928133010864,
|
| 206 |
+
"learning_rate": 9.488837789171455e-05,
|
| 207 |
+
"loss": 4.173,
|
| 208 |
+
"step": 2600
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.08122621500879951,
|
| 212 |
+
"grad_norm": 0.5555790662765503,
|
| 213 |
+
"learning_rate": 9.443843558963266e-05,
|
| 214 |
+
"loss": 4.1537,
|
| 215 |
+
"step": 2700
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.08423459334245875,
|
| 219 |
+
"grad_norm": 0.7323206067085266,
|
| 220 |
+
"learning_rate": 9.397067778366902e-05,
|
| 221 |
+
"loss": 4.1276,
|
| 222 |
+
"step": 2800
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 0.08724297167611798,
|
| 226 |
+
"grad_norm": 0.7370960712432861,
|
| 227 |
+
"learning_rate": 9.348529199934959e-05,
|
| 228 |
+
"loss": 4.1065,
|
| 229 |
+
"step": 2900
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 0.09025135000977723,
|
| 233 |
+
"grad_norm": 0.7088249325752258,
|
| 234 |
+
"learning_rate": 9.298247282931048e-05,
|
| 235 |
+
"loss": 4.0822,
|
| 236 |
+
"step": 3000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.09241738241001188,
|
| 240 |
+
"eval_loss": 4.063363075256348,
|
| 241 |
+
"eval_runtime": 17.3504,
|
| 242 |
+
"eval_samples_per_second": 244.259,
|
| 243 |
+
"eval_steps_per_second": 15.273,
|
| 244 |
+
"step": 3072
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"epoch": 0.09325972834343647,
|
| 248 |
+
"grad_norm": 0.6703213453292847,
|
| 249 |
+
"learning_rate": 9.246242185528508e-05,
|
| 250 |
+
"loss": 4.071,
|
| 251 |
+
"step": 3100
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 0.0962681066770957,
|
| 255 |
+
"grad_norm": 0.6293588876724243,
|
| 256 |
+
"learning_rate": 9.192534756728936e-05,
|
| 257 |
+
"loss": 4.0473,
|
| 258 |
+
"step": 3200
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 0.09927648501075495,
|
| 262 |
+
"grad_norm": 0.5394164323806763,
|
| 263 |
+
"learning_rate": 9.137146528003762e-05,
|
| 264 |
+
"loss": 4.0297,
|
| 265 |
+
"step": 3300
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.1022848633444142,
|
| 269 |
+
"grad_norm": 0.6107392311096191,
|
| 270 |
+
"learning_rate": 9.080099704662214e-05,
|
| 271 |
+
"loss": 4.0133,
|
| 272 |
+
"step": 3400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.10529324167807344,
|
| 276 |
+
"grad_norm": 0.5527825355529785,
|
| 277 |
+
"learning_rate": 9.021417156949143e-05,
|
| 278 |
+
"loss": 3.9874,
|
| 279 |
+
"step": 3500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.10830162001173267,
|
| 283 |
+
"grad_norm": 0.579561173915863,
|
| 284 |
+
"learning_rate": 8.961122410876268e-05,
|
| 285 |
+
"loss": 3.9776,
|
| 286 |
+
"step": 3600
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.11130999834539192,
|
| 290 |
+
"grad_norm": 0.5448581576347351,
|
| 291 |
+
"learning_rate": 8.899239638790535e-05,
|
| 292 |
+
"loss": 3.9607,
|
| 293 |
+
"step": 3700
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.11431837667905116,
|
| 297 |
+
"grad_norm": 0.5136449337005615,
|
| 298 |
+
"learning_rate": 8.835793649683342e-05,
|
| 299 |
+
"loss": 3.9548,
|
| 300 |
+
"step": 3800
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.1173267550127104,
|
| 304 |
+
"grad_norm": 0.5693534016609192,
|
| 305 |
+
"learning_rate": 8.770809879244541e-05,
|
| 306 |
+
"loss": 3.9344,
|
| 307 |
+
"step": 3900
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.12033513334636964,
|
| 311 |
+
"grad_norm": 0.6340081095695496,
|
| 312 |
+
"learning_rate": 8.704314379665195e-05,
|
| 313 |
+
"loss": 3.9257,
|
| 314 |
+
"step": 4000
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.12322317654668251,
|
| 318 |
+
"eval_loss": 3.9077117443084717,
|
| 319 |
+
"eval_runtime": 17.3823,
|
| 320 |
+
"eval_samples_per_second": 243.811,
|
| 321 |
+
"eval_steps_per_second": 15.245,
|
| 322 |
+
"step": 4096
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.12334351168002888,
|
| 326 |
+
"grad_norm": 0.5064478516578674,
|
| 327 |
+
"learning_rate": 8.636333809193176e-05,
|
| 328 |
+
"loss": 3.9101,
|
| 329 |
+
"step": 4100
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.12635189001368813,
|
| 333 |
+
"grad_norm": 0.5670201182365417,
|
| 334 |
+
"learning_rate": 8.566895421445784e-05,
|
| 335 |
+
"loss": 3.9048,
|
| 336 |
+
"step": 4200
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.12936026834734737,
|
| 340 |
+
"grad_norm": 0.5314868092536926,
|
| 341 |
+
"learning_rate": 8.496027054483702e-05,
|
| 342 |
+
"loss": 3.892,
|
| 343 |
+
"step": 4300
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 0.1323686466810066,
|
| 347 |
+
"grad_norm": 0.4874283969402313,
|
| 348 |
+
"learning_rate": 8.423757119650619e-05,
|
| 349 |
+
"loss": 3.8769,
|
| 350 |
+
"step": 4400
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.13537702501466584,
|
| 354 |
+
"grad_norm": 0.5154730677604675,
|
| 355 |
+
"learning_rate": 8.35011459018305e-05,
|
| 356 |
+
"loss": 3.8604,
|
| 357 |
+
"step": 4500
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 0.13838540334832508,
|
| 361 |
+
"grad_norm": 0.5159196853637695,
|
| 362 |
+
"learning_rate": 8.275128989594867e-05,
|
| 363 |
+
"loss": 3.8603,
|
| 364 |
+
"step": 4600
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.14139378168198433,
|
| 368 |
+
"grad_norm": 0.507743775844574,
|
| 369 |
+
"learning_rate": 8.198830379841233e-05,
|
| 370 |
+
"loss": 3.8484,
|
| 371 |
+
"step": 4700
|
| 372 |
+
},
|
| 373 |
+
{
|
| 374 |
+
"epoch": 0.14440216001564357,
|
| 375 |
+
"grad_norm": 0.5081307888031006,
|
| 376 |
+
"learning_rate": 8.121249349266684e-05,
|
| 377 |
+
"loss": 3.8355,
|
| 378 |
+
"step": 4800
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.14741053834930282,
|
| 382 |
+
"grad_norm": 0.5002135634422302,
|
| 383 |
+
"learning_rate": 8.042417000342151e-05,
|
| 384 |
+
"loss": 3.8253,
|
| 385 |
+
"step": 4900
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.15041891668296206,
|
| 389 |
+
"grad_norm": 0.45690950751304626,
|
| 390 |
+
"learning_rate": 7.962364937195902e-05,
|
| 391 |
+
"loss": 3.8184,
|
| 392 |
+
"step": 5000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.15342729501662128,
|
| 396 |
+
"grad_norm": 0.5714775323867798,
|
| 397 |
+
"learning_rate": 7.88112525294334e-05,
|
| 398 |
+
"loss": 3.8074,
|
| 399 |
+
"step": 5100
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 0.15402897068335314,
|
| 403 |
+
"eval_loss": 3.8044040203094482,
|
| 404 |
+
"eval_runtime": 17.3399,
|
| 405 |
+
"eval_samples_per_second": 244.408,
|
| 406 |
+
"eval_steps_per_second": 15.283,
|
| 407 |
+
"step": 5120
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.15643567335028052,
|
| 411 |
+
"grad_norm": 0.530471920967102,
|
| 412 |
+
"learning_rate": 7.798730516820779e-05,
|
| 413 |
+
"loss": 3.8047,
|
| 414 |
+
"step": 5200
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 0.15944405168393977,
|
| 418 |
+
"grad_norm": 0.5580628514289856,
|
| 419 |
+
"learning_rate": 7.715213761128337e-05,
|
| 420 |
+
"loss": 3.7919,
|
| 421 |
+
"step": 5300
|
| 422 |
+
},
|
| 423 |
+
{
|
| 424 |
+
"epoch": 0.16245243001759901,
|
| 425 |
+
"grad_norm": 0.5446878671646118,
|
| 426 |
+
"learning_rate": 7.63060846798719e-05,
|
| 427 |
+
"loss": 3.7852,
|
| 428 |
+
"step": 5400
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.16546080835125826,
|
| 432 |
+
"grad_norm": 0.5348371863365173,
|
| 433 |
+
"learning_rate": 7.544948555916485e-05,
|
| 434 |
+
"loss": 3.781,
|
| 435 |
+
"step": 5500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 0.1684691866849175,
|
| 439 |
+
"grad_norm": 0.4923047721385956,
|
| 440 |
+
"learning_rate": 7.4582683662353e-05,
|
| 441 |
+
"loss": 3.7734,
|
| 442 |
+
"step": 5600
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.17147756501857675,
|
| 446 |
+
"grad_norm": 0.4678339958190918,
|
| 447 |
+
"learning_rate": 7.370602649295112e-05,
|
| 448 |
+
"loss": 3.7631,
|
| 449 |
+
"step": 5700
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.17448594335223597,
|
| 453 |
+
"grad_norm": 0.5229213237762451,
|
| 454 |
+
"learning_rate": 7.281986550548268e-05,
|
| 455 |
+
"loss": 3.7529,
|
| 456 |
+
"step": 5800
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.1774943216858952,
|
| 460 |
+
"grad_norm": 0.6075284481048584,
|
| 461 |
+
"learning_rate": 7.192455596458073e-05,
|
| 462 |
+
"loss": 3.742,
|
| 463 |
+
"step": 5900
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.18050270001955446,
|
| 467 |
+
"grad_norm": 0.5195209383964539,
|
| 468 |
+
"learning_rate": 7.102045680256113e-05,
|
| 469 |
+
"loss": 3.7437,
|
| 470 |
+
"step": 6000
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.1835110783532137,
|
| 474 |
+
"grad_norm": 0.5067561864852905,
|
| 475 |
+
"learning_rate": 7.010793047552565e-05,
|
| 476 |
+
"loss": 3.7379,
|
| 477 |
+
"step": 6100
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.18483476482002376,
|
| 481 |
+
"eval_loss": 3.7293174266815186,
|
| 482 |
+
"eval_runtime": 17.355,
|
| 483 |
+
"eval_samples_per_second": 244.194,
|
| 484 |
+
"eval_steps_per_second": 15.269,
|
| 485 |
+
"step": 6144
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.18651945668687295,
|
| 489 |
+
"grad_norm": 0.46130332350730896,
|
| 490 |
+
"learning_rate": 6.918734281805204e-05,
|
| 491 |
+
"loss": 3.73,
|
| 492 |
+
"step": 6200
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.1895278350205322,
|
| 496 |
+
"grad_norm": 0.4464566707611084,
|
| 497 |
+
"learning_rate": 6.825906289652998e-05,
|
| 498 |
+
"loss": 3.7287,
|
| 499 |
+
"step": 6300
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.1925362133541914,
|
| 503 |
+
"grad_norm": 0.43425479531288147,
|
| 504 |
+
"learning_rate": 6.732346286120116e-05,
|
| 505 |
+
"loss": 3.7126,
|
| 506 |
+
"step": 6400
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.19554459168785066,
|
| 510 |
+
"grad_norm": 0.4764977693557739,
|
| 511 |
+
"learning_rate": 6.638091779696314e-05,
|
| 512 |
+
"loss": 3.7079,
|
| 513 |
+
"step": 6500
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.1985529700215099,
|
| 517 |
+
"grad_norm": 0.5161123275756836,
|
| 518 |
+
"learning_rate": 6.543180557299673e-05,
|
| 519 |
+
"loss": 3.703,
|
| 520 |
+
"step": 6600
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.20156134835516915,
|
| 524 |
+
"grad_norm": 0.4808368384838104,
|
| 525 |
+
"learning_rate": 6.447650669127692e-05,
|
| 526 |
+
"loss": 3.7042,
|
| 527 |
+
"step": 6700
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.2045697266888284,
|
| 531 |
+
"grad_norm": 0.48772090673446655,
|
| 532 |
+
"learning_rate": 6.35154041340287e-05,
|
| 533 |
+
"loss": 3.7031,
|
| 534 |
+
"step": 6800
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.20757810502248764,
|
| 538 |
+
"grad_norm": 0.4356532096862793,
|
| 539 |
+
"learning_rate": 6.254888321018817e-05,
|
| 540 |
+
"loss": 3.6848,
|
| 541 |
+
"step": 6900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.21058648335614688,
|
| 545 |
+
"grad_norm": 0.4405314028263092,
|
| 546 |
+
"learning_rate": 6.157733140093107e-05,
|
| 547 |
+
"loss": 3.6811,
|
| 548 |
+
"step": 7000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.2135948616898061,
|
| 552 |
+
"grad_norm": 0.46900585293769836,
|
| 553 |
+
"learning_rate": 6.060113820433041e-05,
|
| 554 |
+
"loss": 3.6719,
|
| 555 |
+
"step": 7100
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.2156405589566944,
|
| 559 |
+
"eval_loss": 3.674034595489502,
|
| 560 |
+
"eval_runtime": 17.3459,
|
| 561 |
+
"eval_samples_per_second": 244.322,
|
| 562 |
+
"eval_steps_per_second": 15.277,
|
| 563 |
+
"step": 7168
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.21660324002346534,
|
| 567 |
+
"grad_norm": 0.4276421368122101,
|
| 568 |
+
"learning_rate": 5.962069497920546e-05,
|
| 569 |
+
"loss": 3.6764,
|
| 570 |
+
"step": 7200
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.2196116183571246,
|
| 574 |
+
"grad_norm": 0.46885746717453003,
|
| 575 |
+
"learning_rate": 5.863639478822488e-05,
|
| 576 |
+
"loss": 3.6707,
|
| 577 |
+
"step": 7300
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.22261999669078383,
|
| 581 |
+
"grad_norm": 0.46067696809768677,
|
| 582 |
+
"learning_rate": 5.764863224032663e-05,
|
| 583 |
+
"loss": 3.6606,
|
| 584 |
+
"step": 7400
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.22562837502444308,
|
| 588 |
+
"grad_norm": 0.4547005295753479,
|
| 589 |
+
"learning_rate": 5.66578033325181e-05,
|
| 590 |
+
"loss": 3.6531,
|
| 591 |
+
"step": 7500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.22863675335810232,
|
| 595 |
+
"grad_norm": 0.4294310212135315,
|
| 596 |
+
"learning_rate": 5.566430529111969e-05,
|
| 597 |
+
"loss": 3.6599,
|
| 598 |
+
"step": 7600
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.23164513169176157,
|
| 602 |
+
"grad_norm": 0.4465465545654297,
|
| 603 |
+
"learning_rate": 5.4668536412515534e-05,
|
| 604 |
+
"loss": 3.6551,
|
| 605 |
+
"step": 7700
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.2346535100254208,
|
| 609 |
+
"grad_norm": 0.4661228060722351,
|
| 610 |
+
"learning_rate": 5.367089590347535e-05,
|
| 611 |
+
"loss": 3.6517,
|
| 612 |
+
"step": 7800
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.23766188835908003,
|
| 616 |
+
"grad_norm": 0.4832689166069031,
|
| 617 |
+
"learning_rate": 5.2671783721111145e-05,
|
| 618 |
+
"loss": 3.639,
|
| 619 |
+
"step": 7900
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.24067026669273928,
|
| 623 |
+
"grad_norm": 0.4388842284679413,
|
| 624 |
+
"learning_rate": 5.1671600412533284e-05,
|
| 625 |
+
"loss": 3.6392,
|
| 626 |
+
"step": 8000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.24367864502639852,
|
| 630 |
+
"grad_norm": 0.46607160568237305,
|
| 631 |
+
"learning_rate": 5.067074695426984e-05,
|
| 632 |
+
"loss": 3.6267,
|
| 633 |
+
"step": 8100
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.24644635309336502,
|
| 637 |
+
"eval_loss": 3.629561185836792,
|
| 638 |
+
"eval_runtime": 17.3661,
|
| 639 |
+
"eval_samples_per_second": 244.039,
|
| 640 |
+
"eval_steps_per_second": 15.26,
|
| 641 |
+
"step": 8192
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.24668702336005777,
|
| 645 |
+
"grad_norm": 0.48243895173072815,
|
| 646 |
+
"learning_rate": 4.966962459151393e-05,
|
| 647 |
+
"loss": 3.6274,
|
| 648 |
+
"step": 8200
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.249695401693717,
|
| 652 |
+
"grad_norm": 0.4598943293094635,
|
| 653 |
+
"learning_rate": 4.866863467726335e-05,
|
| 654 |
+
"loss": 3.6255,
|
| 655 |
+
"step": 8300
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.25270378002737626,
|
| 659 |
+
"grad_norm": 1.1347087621688843,
|
| 660 |
+
"learning_rate": 4.7668178511416815e-05,
|
| 661 |
+
"loss": 3.622,
|
| 662 |
+
"step": 8400
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.2557121583610355,
|
| 666 |
+
"grad_norm": 0.45290637016296387,
|
| 667 |
+
"learning_rate": 4.666865717989178e-05,
|
| 668 |
+
"loss": 3.6162,
|
| 669 |
+
"step": 8500
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.25872053669469475,
|
| 673 |
+
"grad_norm": 0.4339381754398346,
|
| 674 |
+
"learning_rate": 4.567047139382766e-05,
|
| 675 |
+
"loss": 3.6235,
|
| 676 |
+
"step": 8600
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.261728915028354,
|
| 680 |
+
"grad_norm": 0.43384066224098206,
|
| 681 |
+
"learning_rate": 4.4674021328939746e-05,
|
| 682 |
+
"loss": 3.6107,
|
| 683 |
+
"step": 8700
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.2647372933620132,
|
| 687 |
+
"grad_norm": 0.4384450912475586,
|
| 688 |
+
"learning_rate": 4.3679706465087344e-05,
|
| 689 |
+
"loss": 3.611,
|
| 690 |
+
"step": 8800
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.2677456716956724,
|
| 694 |
+
"grad_norm": 0.41916537284851074,
|
| 695 |
+
"learning_rate": 4.268792542612125e-05,
|
| 696 |
+
"loss": 3.6025,
|
| 697 |
+
"step": 8900
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.27075405002933167,
|
| 701 |
+
"grad_norm": 0.41136473417282104,
|
| 702 |
+
"learning_rate": 4.169907582007412e-05,
|
| 703 |
+
"loss": 3.6052,
|
| 704 |
+
"step": 9000
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.2737624283629909,
|
| 708 |
+
"grad_norm": 0.4844518005847931,
|
| 709 |
+
"learning_rate": 4.071355407975825e-05,
|
| 710 |
+
"loss": 3.5891,
|
| 711 |
+
"step": 9100
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.27677080669665016,
|
| 715 |
+
"grad_norm": 0.41130563616752625,
|
| 716 |
+
"learning_rate": 3.9731755303834453e-05,
|
| 717 |
+
"loss": 3.5901,
|
| 718 |
+
"step": 9200
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.27725214723003566,
|
| 722 |
+
"eval_loss": 3.5964441299438477,
|
| 723 |
+
"eval_runtime": 17.3427,
|
| 724 |
+
"eval_samples_per_second": 244.368,
|
| 725 |
+
"eval_steps_per_second": 15.28,
|
| 726 |
+
"step": 9216
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.2797791850303094,
|
| 730 |
+
"grad_norm": 0.4459775686264038,
|
| 731 |
+
"learning_rate": 3.875407309841582e-05,
|
| 732 |
+
"loss": 3.5833,
|
| 733 |
+
"step": 9300
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 0.28278756336396865,
|
| 737 |
+
"grad_norm": 0.463514119386673,
|
| 738 |
+
"learning_rate": 3.778089941926976e-05,
|
| 739 |
+
"loss": 3.5885,
|
| 740 |
+
"step": 9400
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.2857959416976279,
|
| 744 |
+
"grad_norm": 0.44732800126075745,
|
| 745 |
+
"learning_rate": 3.681262441468189e-05,
|
| 746 |
+
"loss": 3.5791,
|
| 747 |
+
"step": 9500
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 0.28880432003128714,
|
| 751 |
+
"grad_norm": 0.4401909112930298,
|
| 752 |
+
"learning_rate": 3.584963626904423e-05,
|
| 753 |
+
"loss": 3.5907,
|
| 754 |
+
"step": 9600
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.2918126983649464,
|
| 758 |
+
"grad_norm": 0.4163687229156494,
|
| 759 |
+
"learning_rate": 3.4892321047231e-05,
|
| 760 |
+
"loss": 3.5757,
|
| 761 |
+
"step": 9700
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.29482107669860563,
|
| 765 |
+
"grad_norm": 0.4307982921600342,
|
| 766 |
+
"learning_rate": 3.3941062539824027e-05,
|
| 767 |
+
"loss": 3.5781,
|
| 768 |
+
"step": 9800
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.2978294550322649,
|
| 772 |
+
"grad_norm": 0.45625069737434387,
|
| 773 |
+
"learning_rate": 3.299624210924994e-05,
|
| 774 |
+
"loss": 3.5735,
|
| 775 |
+
"step": 9900
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.3008378333659241,
|
| 779 |
+
"grad_norm": 0.45930513739585876,
|
| 780 |
+
"learning_rate": 3.205823853689076e-05,
|
| 781 |
+
"loss": 3.5721,
|
| 782 |
+
"step": 10000
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.3038462116995833,
|
| 786 |
+
"grad_norm": 0.45670804381370544,
|
| 787 |
+
"learning_rate": 3.112742787122936e-05,
|
| 788 |
+
"loss": 3.5775,
|
| 789 |
+
"step": 10100
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 0.30685459003324256,
|
| 793 |
+
"grad_norm": 0.46425703167915344,
|
| 794 |
+
"learning_rate": 3.020418327709045e-05,
|
| 795 |
+
"loss": 3.5696,
|
| 796 |
+
"step": 10200
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 0.3080579413667063,
|
| 800 |
+
"eval_loss": 3.5699303150177,
|
| 801 |
+
"eval_runtime": 17.3654,
|
| 802 |
+
"eval_samples_per_second": 244.049,
|
| 803 |
+
"eval_steps_per_second": 15.26,
|
| 804 |
+
"step": 10240
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 0.3098629683669018,
|
| 808 |
+
"grad_norm": 0.43032941222190857,
|
| 809 |
+
"learning_rate": 2.928887488603769e-05,
|
| 810 |
+
"loss": 3.5676,
|
| 811 |
+
"step": 10300
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 0.31287134670056105,
|
| 815 |
+
"grad_norm": 0.43331944942474365,
|
| 816 |
+
"learning_rate": 2.8381869647986746e-05,
|
| 817 |
+
"loss": 3.5622,
|
| 818 |
+
"step": 10400
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.3158797250342203,
|
| 822 |
+
"grad_norm": 0.42698901891708374,
|
| 823 |
+
"learning_rate": 2.7483531184094124e-05,
|
| 824 |
+
"loss": 3.5639,
|
| 825 |
+
"step": 10500
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.31888810336787954,
|
| 829 |
+
"grad_norm": 0.4351001977920532,
|
| 830 |
+
"learning_rate": 2.659421964098008e-05,
|
| 831 |
+
"loss": 3.562,
|
| 832 |
+
"step": 10600
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 0.3218964817015388,
|
| 836 |
+
"grad_norm": 0.43403735756874084,
|
| 837 |
+
"learning_rate": 2.5714291546345044e-05,
|
| 838 |
+
"loss": 3.5602,
|
| 839 |
+
"step": 10700
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 0.32490486003519803,
|
| 843 |
+
"grad_norm": 0.4507596790790558,
|
| 844 |
+
"learning_rate": 2.484409966603632e-05,
|
| 845 |
+
"loss": 3.5584,
|
| 846 |
+
"step": 10800
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 0.3279132383688573,
|
| 850 |
+
"grad_norm": 0.4281117022037506,
|
| 851 |
+
"learning_rate": 2.3983992862623362e-05,
|
| 852 |
+
"loss": 3.5482,
|
| 853 |
+
"step": 10900
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 0.3309216167025165,
|
| 857 |
+
"grad_norm": 0.43798819184303284,
|
| 858 |
+
"learning_rate": 2.313431595553762e-05,
|
| 859 |
+
"loss": 3.5534,
|
| 860 |
+
"step": 11000
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.33392999503617576,
|
| 864 |
+
"grad_norm": 0.41813981533050537,
|
| 865 |
+
"learning_rate": 2.2295409582833425e-05,
|
| 866 |
+
"loss": 3.551,
|
| 867 |
+
"step": 11100
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 0.336938373369835,
|
| 871 |
+
"grad_norm": 0.4235725998878479,
|
| 872 |
+
"learning_rate": 2.1467610064625133e-05,
|
| 873 |
+
"loss": 3.5561,
|
| 874 |
+
"step": 11200
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 0.3388637355033769,
|
| 878 |
+
"eval_loss": 3.550001859664917,
|
| 879 |
+
"eval_runtime": 17.3651,
|
| 880 |
+
"eval_samples_per_second": 244.052,
|
| 881 |
+
"eval_steps_per_second": 15.26,
|
| 882 |
+
"step": 11264
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.33994675170349425,
|
| 886 |
+
"grad_norm": 0.41261082887649536,
|
| 887 |
+
"learning_rate": 2.0651249268255396e-05,
|
| 888 |
+
"loss": 3.5471,
|
| 889 |
+
"step": 11300
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.3429551300371535,
|
| 893 |
+
"grad_norm": 0.3950221836566925,
|
| 894 |
+
"learning_rate": 1.9846654475248443e-05,
|
| 895 |
+
"loss": 3.547,
|
| 896 |
+
"step": 11400
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.3459635083708127,
|
| 900 |
+
"grad_norm": 0.3985987901687622,
|
| 901 |
+
"learning_rate": 1.9054148250102126e-05,
|
| 902 |
+
"loss": 3.5484,
|
| 903 |
+
"step": 11500
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.34897188670447193,
|
| 907 |
+
"grad_norm": 0.411224901676178,
|
| 908 |
+
"learning_rate": 1.8274048310970558e-05,
|
| 909 |
+
"loss": 3.5431,
|
| 910 |
+
"step": 11600
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.3519802650381312,
|
| 914 |
+
"grad_norm": 0.40982648730278015,
|
| 915 |
+
"learning_rate": 1.7506667402290234e-05,
|
| 916 |
+
"loss": 3.5303,
|
| 917 |
+
"step": 11700
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.3549886433717904,
|
| 921 |
+
"grad_norm": 0.416829377412796,
|
| 922 |
+
"learning_rate": 1.6752313169399547e-05,
|
| 923 |
+
"loss": 3.5363,
|
| 924 |
+
"step": 11800
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.35799702170544967,
|
| 928 |
+
"grad_norm": 0.41838958859443665,
|
| 929 |
+
"learning_rate": 1.601128803520306e-05,
|
| 930 |
+
"loss": 3.5364,
|
| 931 |
+
"step": 11900
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 0.3610054000391089,
|
| 935 |
+
"grad_norm": 0.41204068064689636,
|
| 936 |
+
"learning_rate": 1.5283889078929137e-05,
|
| 937 |
+
"loss": 3.5384,
|
| 938 |
+
"step": 12000
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.36401377837276816,
|
| 942 |
+
"grad_norm": 0.4131161570549011,
|
| 943 |
+
"learning_rate": 1.4570407917030005e-05,
|
| 944 |
+
"loss": 3.537,
|
| 945 |
+
"step": 12100
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.3670221567064274,
|
| 949 |
+
"grad_norm": 0.41995105147361755,
|
| 950 |
+
"learning_rate": 1.387113058627183e-05,
|
| 951 |
+
"loss": 3.5263,
|
| 952 |
+
"step": 12200
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.3696695296400475,
|
| 956 |
+
"eval_loss": 3.5356359481811523,
|
| 957 |
+
"eval_runtime": 17.3613,
|
| 958 |
+
"eval_samples_per_second": 244.107,
|
| 959 |
+
"eval_steps_per_second": 15.264,
|
| 960 |
+
"step": 12288
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.37003053504008665,
|
| 964 |
+
"grad_norm": 0.40922918915748596,
|
| 965 |
+
"learning_rate": 1.3186337429061708e-05,
|
| 966 |
+
"loss": 3.5311,
|
| 967 |
+
"step": 12300
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 0.3730389133737459,
|
| 971 |
+
"grad_norm": 0.4003280997276306,
|
| 972 |
+
"learning_rate": 1.2516302981057498e-05,
|
| 973 |
+
"loss": 3.5335,
|
| 974 |
+
"step": 12400
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.37604729170740514,
|
| 978 |
+
"grad_norm": 0.5036678314208984,
|
| 979 |
+
"learning_rate": 1.186129586110578e-05,
|
| 980 |
+
"loss": 3.5342,
|
| 981 |
+
"step": 12500
|
| 982 |
+
},
|
| 983 |
+
{
|
| 984 |
+
"epoch": 0.3790556700410644,
|
| 985 |
+
"grad_norm": 0.4126069247722626,
|
| 986 |
+
"learning_rate": 1.1221578663551536e-05,
|
| 987 |
+
"loss": 3.531,
|
| 988 |
+
"step": 12600
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 0.38206404837472363,
|
| 992 |
+
"grad_norm": 0.41981491446495056,
|
| 993 |
+
"learning_rate": 1.0597407852963476e-05,
|
| 994 |
+
"loss": 3.525,
|
| 995 |
+
"step": 12700
|
| 996 |
+
},
|
| 997 |
+
{
|
| 998 |
+
"epoch": 0.3850724267083828,
|
| 999 |
+
"grad_norm": 0.38903629779815674,
|
| 1000 |
+
"learning_rate": 9.98903366131651e-06,
|
| 1001 |
+
"loss": 3.5278,
|
| 1002 |
+
"step": 12800
|
| 1003 |
+
},
|
| 1004 |
+
{
|
| 1005 |
+
"epoch": 0.38808080504204207,
|
| 1006 |
+
"grad_norm": 0.4000399708747864,
|
| 1007 |
+
"learning_rate": 9.39669998767294e-06,
|
| 1008 |
+
"loss": 3.5294,
|
| 1009 |
+
"step": 12900
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 0.3910891833757013,
|
| 1013 |
+
"grad_norm": 0.4160057604312897,
|
| 1014 |
+
"learning_rate": 8.820644300402764e-06,
|
| 1015 |
+
"loss": 3.5207,
|
| 1016 |
+
"step": 13000
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.39409756170936056,
|
| 1020 |
+
"grad_norm": 0.4213537275791168,
|
| 1021 |
+
"learning_rate": 8.261097541981678e-06,
|
| 1022 |
+
"loss": 3.524,
|
| 1023 |
+
"step": 13100
|
| 1024 |
+
},
|
| 1025 |
+
{
|
| 1026 |
+
"epoch": 0.3971059400430198,
|
| 1027 |
+
"grad_norm": 0.38205161690711975,
|
| 1028 |
+
"learning_rate": 7.718284036405666e-06,
|
| 1029 |
+
"loss": 3.5204,
|
| 1030 |
+
"step": 13200
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.40011431837667905,
|
| 1034 |
+
"grad_norm": 0.38554006814956665,
|
| 1035 |
+
"learning_rate": 7.192421399258803e-06,
|
| 1036 |
+
"loss": 3.5247,
|
| 1037 |
+
"step": 13300
|
| 1038 |
+
},
|
| 1039 |
+
{
|
| 1040 |
+
"epoch": 0.4004753237767182,
|
| 1041 |
+
"eval_loss": 3.5258023738861084,
|
| 1042 |
+
"eval_runtime": 17.3863,
|
| 1043 |
+
"eval_samples_per_second": 243.756,
|
| 1044 |
+
"eval_steps_per_second": 15.242,
|
| 1045 |
+
"step": 13312
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.4031226967103383,
|
| 1049 |
+
"grad_norm": 0.38247838616371155,
|
| 1050 |
+
"learning_rate": 6.683720450470471e-06,
|
| 1051 |
+
"loss": 3.5247,
|
| 1052 |
+
"step": 13400
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.40613107504399754,
|
| 1056 |
+
"grad_norm": 0.3926681578159332,
|
| 1057 |
+
"learning_rate": 6.192385129797057e-06,
|
| 1058 |
+
"loss": 3.5208,
|
| 1059 |
+
"step": 13500
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.4091394533776568,
|
| 1063 |
+
"grad_norm": 0.4134446382522583,
|
| 1064 |
+
"learning_rate": 5.7186124150618355e-06,
|
| 1065 |
+
"loss": 3.5168,
|
| 1066 |
+
"step": 13600
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.412147831711316,
|
| 1070 |
+
"grad_norm": 0.3735310435295105,
|
| 1071 |
+
"learning_rate": 5.262592243185998e-06,
|
| 1072 |
+
"loss": 3.5276,
|
| 1073 |
+
"step": 13700
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.41515621004497527,
|
| 1077 |
+
"grad_norm": 0.390245258808136,
|
| 1078 |
+
"learning_rate": 4.824507434042369e-06,
|
| 1079 |
+
"loss": 3.5207,
|
| 1080 |
+
"step": 13800
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.4181645883786345,
|
| 1084 |
+
"grad_norm": 0.38336294889450073,
|
| 1085 |
+
"learning_rate": 4.4045336171622985e-06,
|
| 1086 |
+
"loss": 3.522,
|
| 1087 |
+
"step": 13900
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.42117296671229376,
|
| 1091 |
+
"grad_norm": 0.37893325090408325,
|
| 1092 |
+
"learning_rate": 4.002839161325356e-06,
|
| 1093 |
+
"loss": 3.5178,
|
| 1094 |
+
"step": 14000
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.424181345045953,
|
| 1098 |
+
"grad_norm": 0.3802947998046875,
|
| 1099 |
+
"learning_rate": 3.6195851070596586e-06,
|
| 1100 |
+
"loss": 3.5185,
|
| 1101 |
+
"step": 14100
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.4271897233796122,
|
| 1105 |
+
"grad_norm": 0.3909587860107422,
|
| 1106 |
+
"learning_rate": 3.2549251020803307e-06,
|
| 1107 |
+
"loss": 3.5182,
|
| 1108 |
+
"step": 14200
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.43019810171327144,
|
| 1112 |
+
"grad_norm": 0.3848523199558258,
|
| 1113 |
+
"learning_rate": 2.9090053396916172e-06,
|
| 1114 |
+
"loss": 3.521,
|
| 1115 |
+
"step": 14300
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.4312811179133888,
|
| 1119 |
+
"eval_loss": 3.520242929458618,
|
| 1120 |
+
"eval_runtime": 17.3542,
|
| 1121 |
+
"eval_samples_per_second": 244.206,
|
| 1122 |
+
"eval_steps_per_second": 15.27,
|
| 1123 |
+
"step": 14336
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 0.4332064800469307,
|
| 1127 |
+
"grad_norm": 0.38906753063201904,
|
| 1128 |
+
"learning_rate": 2.5819645001775474e-06,
|
| 1129 |
+
"loss": 3.5214,
|
| 1130 |
+
"step": 14400
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.43621485838058993,
|
| 1134 |
+
"grad_norm": 0.37520134449005127,
|
| 1135 |
+
"learning_rate": 2.2739336952046e-06,
|
| 1136 |
+
"loss": 3.5155,
|
| 1137 |
+
"step": 14500
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.4392232367142492,
|
| 1141 |
+
"grad_norm": 0.4010186195373535,
|
| 1142 |
+
"learning_rate": 1.9850364152586344e-06,
|
| 1143 |
+
"loss": 3.5169,
|
| 1144 |
+
"step": 14600
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.4422316150479084,
|
| 1148 |
+
"grad_norm": 0.37698066234588623,
|
| 1149 |
+
"learning_rate": 1.7153884801371756e-06,
|
| 1150 |
+
"loss": 3.5171,
|
| 1151 |
+
"step": 14700
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 0.44523999338156767,
|
| 1155 |
+
"grad_norm": 0.3856140971183777,
|
| 1156 |
+
"learning_rate": 1.4650979925169117e-06,
|
| 1157 |
+
"loss": 3.5133,
|
| 1158 |
+
"step": 14800
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 0.4482483717152269,
|
| 1162 |
+
"grad_norm": 0.41618818044662476,
|
| 1163 |
+
"learning_rate": 1.234265294614978e-06,
|
| 1164 |
+
"loss": 3.5186,
|
| 1165 |
+
"step": 14900
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 0.45125675004888616,
|
| 1169 |
+
"grad_norm": 0.3722720742225647,
|
| 1170 |
+
"learning_rate": 1.0229829279615133e-06,
|
| 1171 |
+
"loss": 3.522,
|
| 1172 |
+
"step": 15000
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.4542651283825454,
|
| 1176 |
+
"grad_norm": 0.376542866230011,
|
| 1177 |
+
"learning_rate": 8.313355962994041e-07,
|
| 1178 |
+
"loss": 3.5216,
|
| 1179 |
+
"step": 15100
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 0.45727350671620465,
|
| 1183 |
+
"grad_norm": 0.36669039726257324,
|
| 1184 |
+
"learning_rate": 6.594001316263587e-07,
|
| 1185 |
+
"loss": 3.5198,
|
| 1186 |
+
"step": 15200
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.4602818850498639,
|
| 1190 |
+
"grad_norm": 0.3666169047355652,
|
| 1191 |
+
"learning_rate": 5.072454633927026e-07,
|
| 1192 |
+
"loss": 3.5164,
|
| 1193 |
+
"step": 15300
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.4620869120500594,
|
| 1197 |
+
"eval_loss": 3.518115520477295,
|
| 1198 |
+
"eval_runtime": 17.3893,
|
| 1199 |
+
"eval_samples_per_second": 243.713,
|
| 1200 |
+
"eval_steps_per_second": 15.239,
|
| 1201 |
+
"step": 15360
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.46329026338352314,
|
| 1205 |
+
"grad_norm": 0.364441454410553,
|
| 1206 |
+
"learning_rate": 3.7493259086733845e-07,
|
| 1207 |
+
"loss": 3.514,
|
| 1208 |
+
"step": 15400
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.4662986417171823,
|
| 1212 |
+
"grad_norm": 0.3645060658454895,
|
| 1213 |
+
"learning_rate": 2.625145586829858e-07,
|
| 1214 |
+
"loss": 3.5173,
|
| 1215 |
+
"step": 15500
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.4693070200508416,
|
| 1219 |
+
"grad_norm": 0.37121596932411194,
|
| 1220 |
+
"learning_rate": 1.7003643557039762e-07,
|
| 1221 |
+
"loss": 3.5157,
|
| 1222 |
+
"step": 15600
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.4723153983845008,
|
| 1226 |
+
"grad_norm": 0.36515143513679504,
|
| 1227 |
+
"learning_rate": 9.753529629017988e-08,
|
| 1228 |
+
"loss": 3.5182,
|
| 1229 |
+
"step": 15700
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.47532377671816006,
|
| 1233 |
+
"grad_norm": 0.36171218752861023,
|
| 1234 |
+
"learning_rate": 4.504020676941445e-08,
|
| 1235 |
+
"loss": 3.5106,
|
| 1236 |
+
"step": 15800
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.4783321550518193,
|
| 1240 |
+
"grad_norm": 0.3873487412929535,
|
| 1241 |
+
"learning_rate": 1.2572212449019072e-08,
|
| 1242 |
+
"loss": 3.5156,
|
| 1243 |
+
"step": 15900
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 0.48134053338547855,
|
| 1247 |
+
"grad_norm": 0.36549946665763855,
|
| 1248 |
+
"learning_rate": 1.4432984660195382e-10,
|
| 1249 |
+
"loss": 3.5122,
|
| 1250 |
+
"step": 16000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.48167145500218106,
|
| 1254 |
+
"step": 16011,
|
| 1255 |
+
"total_flos": 2.6742109246125507e+18,
|
| 1256 |
+
"train_loss": 3.9105984272824914,
|
| 1257 |
+
"train_runtime": 32974.1722,
|
| 1258 |
+
"train_samples_per_second": 62.152,
|
| 1259 |
+
"train_steps_per_second": 0.486
|
| 1260 |
+
}
|
| 1261 |
+
],
|
| 1262 |
+
"logging_steps": 100,
|
| 1263 |
+
"max_steps": 16011,
|
| 1264 |
+
"num_input_tokens_seen": 0,
|
| 1265 |
+
"num_train_epochs": 1,
|
| 1266 |
+
"save_steps": 2048,
|
| 1267 |
+
"stateful_callbacks": {
|
| 1268 |
+
"TrainerControl": {
|
| 1269 |
+
"args": {
|
| 1270 |
+
"should_epoch_stop": false,
|
| 1271 |
+
"should_evaluate": false,
|
| 1272 |
+
"should_log": false,
|
| 1273 |
+
"should_save": true,
|
| 1274 |
+
"should_training_stop": true
|
| 1275 |
+
},
|
| 1276 |
+
"attributes": {}
|
| 1277 |
+
}
|
| 1278 |
+
},
|
| 1279 |
+
"total_flos": 2.6742109246125507e+18,
|
| 1280 |
+
"train_batch_size": 4,
|
| 1281 |
+
"trial_name": null,
|
| 1282 |
+
"trial_params": null
|
| 1283 |
+
}
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ab45769ad5b5d979d988bee72018f6e6933502290a9e3e331dc71df796d4ee54
|
| 3 |
+
size 7544
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/training_eval_loss.png
ADDED
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/training_loss.png
ADDED
|
qian_sg_backup/qwen3/qwen3_p223m_t1_2_suite-owt_scaling_v3_family-qwen3_geometry_id-euclid_variant-base_init_slope-std/attempt3_20260310_211619/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|