Upload 13 files
Browse files- README.md +52 -3
- adapter_config.json +31 -0
- adapter_model.safetensors +3 -0
- all_results.json +14 -0
- config.json +28 -0
- eval_results.json +8 -0
- special_tokens_map.json +24 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +42 -0
- train_results.json +9 -0
- trainer_state.json +2432 -0
- training_args.bin +3 -0
README.md
CHANGED
|
@@ -1,3 +1,52 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
library_name: peft
|
| 4 |
+
tags:
|
| 5 |
+
- alignment-handbook
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
base_model: TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
|
| 8 |
+
datasets:
|
| 9 |
+
- Ritvik19/open-hermes-2_5-reformatted
|
| 10 |
+
model-index:
|
| 11 |
+
- name: openhermes-tinyllama-sft-qlora
|
| 12 |
+
results: []
|
| 13 |
+
---
|
| 14 |
+
|
| 15 |
+
**Note**: This model card has been generated automatically according to the information the Trainer had access to.
|
| 16 |
+
Visit the [model card](https://ritvik19.github.io/small-llms/) to see the full description.
|
| 17 |
+
|
| 18 |
+
# openhermes-tinyllama-sft-qlora
|
| 19 |
+
|
| 20 |
+
This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T](https://huggingface.co/TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T) on the Ritvik19/open-hermes-2_5-reformatted dataset.
|
| 21 |
+
It achieves the following results on the evaluation set:
|
| 22 |
+
- Loss: 1.2783
|
| 23 |
+
|
| 24 |
+
### Training hyperparameters
|
| 25 |
+
|
| 26 |
+
The following hyperparameters were used during training:
|
| 27 |
+
- learning_rate: 0.0002
|
| 28 |
+
- train_batch_size: 1
|
| 29 |
+
- eval_batch_size: 1
|
| 30 |
+
- seed: 42
|
| 31 |
+
- distributed_type: multi-GPU
|
| 32 |
+
- gradient_accumulation_steps: 128
|
| 33 |
+
- total_train_batch_size: 128
|
| 34 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 35 |
+
- lr_scheduler_type: cosine
|
| 36 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 37 |
+
- num_epochs: 1
|
| 38 |
+
|
| 39 |
+
### Training results
|
| 40 |
+
|
| 41 |
+
| Training Loss | Epoch | Step | Validation Loss |
|
| 42 |
+
|:-------------:|:------:|:----:|:---------------:|
|
| 43 |
+
| 1.1143 | 0.9996 | 1708 | 1.2783 |
|
| 44 |
+
|
| 45 |
+
|
| 46 |
+
### Framework versions
|
| 47 |
+
|
| 48 |
+
- PEFT 0.7.1
|
| 49 |
+
- Transformers 4.40.1
|
| 50 |
+
- Pytorch 2.1.2+cu121
|
| 51 |
+
- Datasets 2.19.0
|
| 52 |
+
- Tokenizers 0.19.1
|
adapter_config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"fan_in_fan_out": false,
|
| 7 |
+
"inference_mode": true,
|
| 8 |
+
"init_lora_weights": true,
|
| 9 |
+
"layers_pattern": null,
|
| 10 |
+
"layers_to_transform": null,
|
| 11 |
+
"loftq_config": {},
|
| 12 |
+
"lora_alpha": 16,
|
| 13 |
+
"lora_dropout": 0.05,
|
| 14 |
+
"megatron_config": null,
|
| 15 |
+
"megatron_core": "megatron.core",
|
| 16 |
+
"modules_to_save": null,
|
| 17 |
+
"peft_type": "LORA",
|
| 18 |
+
"r": 16,
|
| 19 |
+
"rank_pattern": {},
|
| 20 |
+
"revision": null,
|
| 21 |
+
"target_modules": [
|
| 22 |
+
"down_proj",
|
| 23 |
+
"v_proj",
|
| 24 |
+
"k_proj",
|
| 25 |
+
"gate_proj",
|
| 26 |
+
"q_proj",
|
| 27 |
+
"up_proj",
|
| 28 |
+
"o_proj"
|
| 29 |
+
],
|
| 30 |
+
"task_type": "CAUSAL_LM"
|
| 31 |
+
}
|
adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:82e98fa6ca46951971efeeb7669132b5989baa206fa0a10e563384cfcbdfb091
|
| 3 |
+
size 25272360
|
all_results.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9995747929973436,
|
| 3 |
+
"eval_loss": 1.278271198272705,
|
| 4 |
+
"eval_runtime": 1254.1335,
|
| 5 |
+
"eval_samples": 23109,
|
| 6 |
+
"eval_samples_per_second": 12.89,
|
| 7 |
+
"eval_steps_per_second": 12.89,
|
| 8 |
+
"total_flos": 2.8130589802160456e+18,
|
| 9 |
+
"train_loss": 0.978924176871637,
|
| 10 |
+
"train_runtime": 53778.1468,
|
| 11 |
+
"train_samples": 1001538,
|
| 12 |
+
"train_samples_per_second": 4.067,
|
| 13 |
+
"train_steps_per_second": 0.032
|
| 14 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"LlamaForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attention_bias": false,
|
| 7 |
+
"attention_dropout": 0.0,
|
| 8 |
+
"bos_token_id": 1,
|
| 9 |
+
"eos_token_id": 2,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 2048,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 5632,
|
| 14 |
+
"max_position_embeddings": 2048,
|
| 15 |
+
"model_type": "llama",
|
| 16 |
+
"num_attention_heads": 32,
|
| 17 |
+
"num_hidden_layers": 22,
|
| 18 |
+
"num_key_value_heads": 4,
|
| 19 |
+
"pretraining_tp": 1,
|
| 20 |
+
"rms_norm_eps": 1e-05,
|
| 21 |
+
"rope_scaling": null,
|
| 22 |
+
"rope_theta": 10000.0,
|
| 23 |
+
"tie_word_embeddings": false,
|
| 24 |
+
"torch_dtype": "bfloat16",
|
| 25 |
+
"transformers_version": "4.40.1",
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 32000
|
| 28 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9995747929973436,
|
| 3 |
+
"eval_loss": 1.278271198272705,
|
| 4 |
+
"eval_runtime": 1254.1335,
|
| 5 |
+
"eval_samples": 23109,
|
| 6 |
+
"eval_samples_per_second": 12.89,
|
| 7 |
+
"eval_steps_per_second": 12.89
|
| 8 |
+
}
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "</s>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<unk>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"0": {
|
| 6 |
+
"content": "<unk>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"1": {
|
| 14 |
+
"content": "<s>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"2": {
|
| 22 |
+
"content": "</s>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
}
|
| 29 |
+
},
|
| 30 |
+
"bos_token": "<s>",
|
| 31 |
+
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
| 32 |
+
"clean_up_tokenization_spaces": false,
|
| 33 |
+
"eos_token": "</s>",
|
| 34 |
+
"legacy": false,
|
| 35 |
+
"model_max_length": 2048,
|
| 36 |
+
"pad_token": "</s>",
|
| 37 |
+
"padding_side": "right",
|
| 38 |
+
"sp_model_kwargs": {},
|
| 39 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 40 |
+
"unk_token": "<unk>",
|
| 41 |
+
"use_default_system_prompt": false
|
| 42 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 0.9995747929973436,
|
| 3 |
+
"total_flos": 2.8130589802160456e+18,
|
| 4 |
+
"train_loss": 0.978924176871637,
|
| 5 |
+
"train_runtime": 53778.1468,
|
| 6 |
+
"train_samples": 1001538,
|
| 7 |
+
"train_samples_per_second": 4.067,
|
| 8 |
+
"train_steps_per_second": 0.032
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2432 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9995747929973436,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1708,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0005852311434410677,
|
| 13 |
+
"grad_norm": 0.515625,
|
| 14 |
+
"learning_rate": 1.1695906432748538e-06,
|
| 15 |
+
"loss": 1.737,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0029261557172053382,
|
| 20 |
+
"grad_norm": 0.5234375,
|
| 21 |
+
"learning_rate": 5.8479532163742686e-06,
|
| 22 |
+
"loss": 1.7512,
|
| 23 |
+
"step": 5
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.0058523114344106765,
|
| 27 |
+
"grad_norm": 0.48046875,
|
| 28 |
+
"learning_rate": 1.1695906432748537e-05,
|
| 29 |
+
"loss": 1.7378,
|
| 30 |
+
"step": 10
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.008778467151616015,
|
| 34 |
+
"grad_norm": 0.474609375,
|
| 35 |
+
"learning_rate": 1.7543859649122806e-05,
|
| 36 |
+
"loss": 1.7494,
|
| 37 |
+
"step": 15
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.011704622868821353,
|
| 41 |
+
"grad_norm": 0.4765625,
|
| 42 |
+
"learning_rate": 2.3391812865497074e-05,
|
| 43 |
+
"loss": 1.7101,
|
| 44 |
+
"step": 20
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.014630778586026691,
|
| 48 |
+
"grad_norm": 0.482421875,
|
| 49 |
+
"learning_rate": 2.9239766081871346e-05,
|
| 50 |
+
"loss": 1.6696,
|
| 51 |
+
"step": 25
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.01755693430323203,
|
| 55 |
+
"grad_norm": 0.34765625,
|
| 56 |
+
"learning_rate": 3.508771929824561e-05,
|
| 57 |
+
"loss": 1.6369,
|
| 58 |
+
"step": 30
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.02048309002043737,
|
| 62 |
+
"grad_norm": 0.287109375,
|
| 63 |
+
"learning_rate": 4.093567251461988e-05,
|
| 64 |
+
"loss": 1.6058,
|
| 65 |
+
"step": 35
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.023409245737642706,
|
| 69 |
+
"grad_norm": 0.2333984375,
|
| 70 |
+
"learning_rate": 4.678362573099415e-05,
|
| 71 |
+
"loss": 1.5587,
|
| 72 |
+
"step": 40
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.026335401454848046,
|
| 76 |
+
"grad_norm": 0.1708984375,
|
| 77 |
+
"learning_rate": 5.2631578947368424e-05,
|
| 78 |
+
"loss": 1.5149,
|
| 79 |
+
"step": 45
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.029261557172053382,
|
| 83 |
+
"grad_norm": 0.140625,
|
| 84 |
+
"learning_rate": 5.847953216374269e-05,
|
| 85 |
+
"loss": 1.4769,
|
| 86 |
+
"step": 50
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.03218771288925872,
|
| 90 |
+
"grad_norm": 0.12060546875,
|
| 91 |
+
"learning_rate": 6.432748538011695e-05,
|
| 92 |
+
"loss": 1.4573,
|
| 93 |
+
"step": 55
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.03511386860646406,
|
| 97 |
+
"grad_norm": 0.09228515625,
|
| 98 |
+
"learning_rate": 7.017543859649122e-05,
|
| 99 |
+
"loss": 1.4219,
|
| 100 |
+
"step": 60
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.0380400243236694,
|
| 104 |
+
"grad_norm": 0.09375,
|
| 105 |
+
"learning_rate": 7.602339181286549e-05,
|
| 106 |
+
"loss": 1.398,
|
| 107 |
+
"step": 65
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.04096618004087474,
|
| 111 |
+
"grad_norm": 0.06884765625,
|
| 112 |
+
"learning_rate": 8.187134502923976e-05,
|
| 113 |
+
"loss": 1.3912,
|
| 114 |
+
"step": 70
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.043892335758080075,
|
| 118 |
+
"grad_norm": 0.06787109375,
|
| 119 |
+
"learning_rate": 8.771929824561403e-05,
|
| 120 |
+
"loss": 1.3633,
|
| 121 |
+
"step": 75
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.04681849147528541,
|
| 125 |
+
"grad_norm": 0.06103515625,
|
| 126 |
+
"learning_rate": 9.35672514619883e-05,
|
| 127 |
+
"loss": 1.3546,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.049744647192490755,
|
| 132 |
+
"grad_norm": 0.05712890625,
|
| 133 |
+
"learning_rate": 9.941520467836257e-05,
|
| 134 |
+
"loss": 1.3309,
|
| 135 |
+
"step": 85
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.05267080290969609,
|
| 139 |
+
"grad_norm": 0.05517578125,
|
| 140 |
+
"learning_rate": 0.00010526315789473685,
|
| 141 |
+
"loss": 1.3234,
|
| 142 |
+
"step": 90
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.05559695862690143,
|
| 146 |
+
"grad_norm": 0.048828125,
|
| 147 |
+
"learning_rate": 0.00011111111111111112,
|
| 148 |
+
"loss": 1.3061,
|
| 149 |
+
"step": 95
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.058523114344106765,
|
| 153 |
+
"grad_norm": 0.048828125,
|
| 154 |
+
"learning_rate": 0.00011695906432748539,
|
| 155 |
+
"loss": 1.2844,
|
| 156 |
+
"step": 100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.06144927006131211,
|
| 160 |
+
"grad_norm": 0.04833984375,
|
| 161 |
+
"learning_rate": 0.00012280701754385965,
|
| 162 |
+
"loss": 1.2973,
|
| 163 |
+
"step": 105
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.06437542577851744,
|
| 167 |
+
"grad_norm": 0.050537109375,
|
| 168 |
+
"learning_rate": 0.0001286549707602339,
|
| 169 |
+
"loss": 1.2836,
|
| 170 |
+
"step": 110
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.06730158149572278,
|
| 174 |
+
"grad_norm": 0.048095703125,
|
| 175 |
+
"learning_rate": 0.0001345029239766082,
|
| 176 |
+
"loss": 1.2723,
|
| 177 |
+
"step": 115
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.07022773721292812,
|
| 181 |
+
"grad_norm": 0.046630859375,
|
| 182 |
+
"learning_rate": 0.00014035087719298245,
|
| 183 |
+
"loss": 1.2634,
|
| 184 |
+
"step": 120
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.07315389293013345,
|
| 188 |
+
"grad_norm": 0.05078125,
|
| 189 |
+
"learning_rate": 0.00014619883040935673,
|
| 190 |
+
"loss": 1.2355,
|
| 191 |
+
"step": 125
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.0760800486473388,
|
| 195 |
+
"grad_norm": 0.05029296875,
|
| 196 |
+
"learning_rate": 0.00015204678362573098,
|
| 197 |
+
"loss": 1.2494,
|
| 198 |
+
"step": 130
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.07900620436454414,
|
| 202 |
+
"grad_norm": 0.054931640625,
|
| 203 |
+
"learning_rate": 0.00015789473684210527,
|
| 204 |
+
"loss": 1.253,
|
| 205 |
+
"step": 135
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.08193236008174948,
|
| 209 |
+
"grad_norm": 0.05322265625,
|
| 210 |
+
"learning_rate": 0.00016374269005847952,
|
| 211 |
+
"loss": 1.2499,
|
| 212 |
+
"step": 140
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.08485851579895481,
|
| 216 |
+
"grad_norm": 0.04833984375,
|
| 217 |
+
"learning_rate": 0.0001695906432748538,
|
| 218 |
+
"loss": 1.2193,
|
| 219 |
+
"step": 145
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.08778467151616015,
|
| 223 |
+
"grad_norm": 0.05712890625,
|
| 224 |
+
"learning_rate": 0.00017543859649122806,
|
| 225 |
+
"loss": 1.2339,
|
| 226 |
+
"step": 150
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.09071082723336549,
|
| 230 |
+
"grad_norm": 0.056884765625,
|
| 231 |
+
"learning_rate": 0.00018128654970760234,
|
| 232 |
+
"loss": 1.2427,
|
| 233 |
+
"step": 155
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.09363698295057082,
|
| 237 |
+
"grad_norm": 0.052490234375,
|
| 238 |
+
"learning_rate": 0.0001871345029239766,
|
| 239 |
+
"loss": 1.2184,
|
| 240 |
+
"step": 160
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.09656313866777616,
|
| 244 |
+
"grad_norm": 0.06298828125,
|
| 245 |
+
"learning_rate": 0.00019298245614035088,
|
| 246 |
+
"loss": 1.2306,
|
| 247 |
+
"step": 165
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.09948929438498151,
|
| 251 |
+
"grad_norm": 0.06298828125,
|
| 252 |
+
"learning_rate": 0.00019883040935672513,
|
| 253 |
+
"loss": 1.214,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.10241545010218685,
|
| 258 |
+
"grad_norm": 0.0673828125,
|
| 259 |
+
"learning_rate": 0.00019999665774502696,
|
| 260 |
+
"loss": 1.2176,
|
| 261 |
+
"step": 175
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.10534160581939218,
|
| 265 |
+
"grad_norm": 0.0634765625,
|
| 266 |
+
"learning_rate": 0.0001999830802170989,
|
| 267 |
+
"loss": 1.204,
|
| 268 |
+
"step": 180
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.10826776153659752,
|
| 272 |
+
"grad_norm": 0.06298828125,
|
| 273 |
+
"learning_rate": 0.00019995905994229593,
|
| 274 |
+
"loss": 1.2153,
|
| 275 |
+
"step": 185
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.11119391725380286,
|
| 279 |
+
"grad_norm": 0.07080078125,
|
| 280 |
+
"learning_rate": 0.00019992459942941906,
|
| 281 |
+
"loss": 1.1936,
|
| 282 |
+
"step": 190
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.1141200729710082,
|
| 286 |
+
"grad_norm": 0.064453125,
|
| 287 |
+
"learning_rate": 0.00019987970227770135,
|
| 288 |
+
"loss": 1.1987,
|
| 289 |
+
"step": 195
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.11704622868821353,
|
| 293 |
+
"grad_norm": 0.07177734375,
|
| 294 |
+
"learning_rate": 0.00019982437317643217,
|
| 295 |
+
"loss": 1.2065,
|
| 296 |
+
"step": 200
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.11997238440541888,
|
| 300 |
+
"grad_norm": 0.0654296875,
|
| 301 |
+
"learning_rate": 0.00019975861790446722,
|
| 302 |
+
"loss": 1.2088,
|
| 303 |
+
"step": 205
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.12289854012262422,
|
| 307 |
+
"grad_norm": 0.07373046875,
|
| 308 |
+
"learning_rate": 0.0001996824433296252,
|
| 309 |
+
"loss": 1.2082,
|
| 310 |
+
"step": 210
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.12582469583982955,
|
| 314 |
+
"grad_norm": 0.07958984375,
|
| 315 |
+
"learning_rate": 0.00019959585740797028,
|
| 316 |
+
"loss": 1.2062,
|
| 317 |
+
"step": 215
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.1287508515570349,
|
| 321 |
+
"grad_norm": 0.0966796875,
|
| 322 |
+
"learning_rate": 0.0001994988691829812,
|
| 323 |
+
"loss": 1.2046,
|
| 324 |
+
"step": 220
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.13167700727424023,
|
| 328 |
+
"grad_norm": 0.07275390625,
|
| 329 |
+
"learning_rate": 0.00019939148878460677,
|
| 330 |
+
"loss": 1.195,
|
| 331 |
+
"step": 225
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.13460316299144556,
|
| 335 |
+
"grad_norm": 0.0859375,
|
| 336 |
+
"learning_rate": 0.00019927372742820779,
|
| 337 |
+
"loss": 1.1807,
|
| 338 |
+
"step": 230
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.1375293187086509,
|
| 342 |
+
"grad_norm": 0.07666015625,
|
| 343 |
+
"learning_rate": 0.0001991455974133857,
|
| 344 |
+
"loss": 1.1887,
|
| 345 |
+
"step": 235
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.14045547442585624,
|
| 349 |
+
"grad_norm": 0.07373046875,
|
| 350 |
+
"learning_rate": 0.0001990071121226979,
|
| 351 |
+
"loss": 1.189,
|
| 352 |
+
"step": 240
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.14338163014306157,
|
| 356 |
+
"grad_norm": 0.0751953125,
|
| 357 |
+
"learning_rate": 0.0001988582860202601,
|
| 358 |
+
"loss": 1.172,
|
| 359 |
+
"step": 245
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.1463077858602669,
|
| 363 |
+
"grad_norm": 0.078125,
|
| 364 |
+
"learning_rate": 0.00019869913465023548,
|
| 365 |
+
"loss": 1.1738,
|
| 366 |
+
"step": 250
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.14923394157747225,
|
| 370 |
+
"grad_norm": 0.08447265625,
|
| 371 |
+
"learning_rate": 0.00019852967463521124,
|
| 372 |
+
"loss": 1.1947,
|
| 373 |
+
"step": 255
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.1521600972946776,
|
| 377 |
+
"grad_norm": 0.080078125,
|
| 378 |
+
"learning_rate": 0.0001983499236744625,
|
| 379 |
+
"loss": 1.1789,
|
| 380 |
+
"step": 260
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.15508625301188295,
|
| 384 |
+
"grad_norm": 0.078125,
|
| 385 |
+
"learning_rate": 0.00019815990054210361,
|
| 386 |
+
"loss": 1.1878,
|
| 387 |
+
"step": 265
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.15801240872908828,
|
| 391 |
+
"grad_norm": 0.072265625,
|
| 392 |
+
"learning_rate": 0.00019795962508512742,
|
| 393 |
+
"loss": 1.1825,
|
| 394 |
+
"step": 270
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.16093856444629362,
|
| 398 |
+
"grad_norm": 0.103515625,
|
| 399 |
+
"learning_rate": 0.00019774911822133216,
|
| 400 |
+
"loss": 1.1848,
|
| 401 |
+
"step": 275
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.16386472016349896,
|
| 405 |
+
"grad_norm": 0.091796875,
|
| 406 |
+
"learning_rate": 0.0001975284019371368,
|
| 407 |
+
"loss": 1.1634,
|
| 408 |
+
"step": 280
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.1667908758807043,
|
| 412 |
+
"grad_norm": 0.078125,
|
| 413 |
+
"learning_rate": 0.0001972974992852847,
|
| 414 |
+
"loss": 1.1539,
|
| 415 |
+
"step": 285
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.16971703159790963,
|
| 419 |
+
"grad_norm": 0.06982421875,
|
| 420 |
+
"learning_rate": 0.00019705643438243584,
|
| 421 |
+
"loss": 1.1656,
|
| 422 |
+
"step": 290
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.17264318731511497,
|
| 426 |
+
"grad_norm": 0.0859375,
|
| 427 |
+
"learning_rate": 0.00019680523240664786,
|
| 428 |
+
"loss": 1.1923,
|
| 429 |
+
"step": 295
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.1755693430323203,
|
| 433 |
+
"grad_norm": 0.07958984375,
|
| 434 |
+
"learning_rate": 0.00019654391959474647,
|
| 435 |
+
"loss": 1.1651,
|
| 436 |
+
"step": 300
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.17849549874952564,
|
| 440 |
+
"grad_norm": 0.08056640625,
|
| 441 |
+
"learning_rate": 0.00019627252323958504,
|
| 442 |
+
"loss": 1.1501,
|
| 443 |
+
"step": 305
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.18142165446673097,
|
| 447 |
+
"grad_norm": 0.0859375,
|
| 448 |
+
"learning_rate": 0.00019599107168719412,
|
| 449 |
+
"loss": 1.1581,
|
| 450 |
+
"step": 310
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.1843478101839363,
|
| 454 |
+
"grad_norm": 0.08349609375,
|
| 455 |
+
"learning_rate": 0.0001956995943338206,
|
| 456 |
+
"loss": 1.1785,
|
| 457 |
+
"step": 315
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.18727396590114165,
|
| 461 |
+
"grad_norm": 0.08447265625,
|
| 462 |
+
"learning_rate": 0.00019539812162285767,
|
| 463 |
+
"loss": 1.1691,
|
| 464 |
+
"step": 320
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.19020012161834698,
|
| 468 |
+
"grad_norm": 0.0751953125,
|
| 469 |
+
"learning_rate": 0.00019508668504166505,
|
| 470 |
+
"loss": 1.1758,
|
| 471 |
+
"step": 325
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.19312627733555232,
|
| 475 |
+
"grad_norm": 0.0703125,
|
| 476 |
+
"learning_rate": 0.00019476531711828027,
|
| 477 |
+
"loss": 1.1582,
|
| 478 |
+
"step": 330
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.19605243305275769,
|
| 482 |
+
"grad_norm": 0.08642578125,
|
| 483 |
+
"learning_rate": 0.0001944340514180212,
|
| 484 |
+
"loss": 1.1767,
|
| 485 |
+
"step": 335
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.19897858876996302,
|
| 489 |
+
"grad_norm": 0.0703125,
|
| 490 |
+
"learning_rate": 0.00019409292253998062,
|
| 491 |
+
"loss": 1.1392,
|
| 492 |
+
"step": 340
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.20190474448716836,
|
| 496 |
+
"grad_norm": 0.1064453125,
|
| 497 |
+
"learning_rate": 0.0001937419661134121,
|
| 498 |
+
"loss": 1.1626,
|
| 499 |
+
"step": 345
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.2048309002043737,
|
| 503 |
+
"grad_norm": 0.08203125,
|
| 504 |
+
"learning_rate": 0.00019338121879400896,
|
| 505 |
+
"loss": 1.1551,
|
| 506 |
+
"step": 350
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.20775705592157903,
|
| 510 |
+
"grad_norm": 0.080078125,
|
| 511 |
+
"learning_rate": 0.00019301071826007576,
|
| 512 |
+
"loss": 1.1495,
|
| 513 |
+
"step": 355
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.21068321163878437,
|
| 517 |
+
"grad_norm": 0.0810546875,
|
| 518 |
+
"learning_rate": 0.00019263050320859283,
|
| 519 |
+
"loss": 1.1514,
|
| 520 |
+
"step": 360
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.2136093673559897,
|
| 524 |
+
"grad_norm": 0.08935546875,
|
| 525 |
+
"learning_rate": 0.00019224061335117472,
|
| 526 |
+
"loss": 1.1649,
|
| 527 |
+
"step": 365
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.21653552307319504,
|
| 531 |
+
"grad_norm": 0.08837890625,
|
| 532 |
+
"learning_rate": 0.0001918410894099224,
|
| 533 |
+
"loss": 1.1433,
|
| 534 |
+
"step": 370
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.21946167879040038,
|
| 538 |
+
"grad_norm": 0.09423828125,
|
| 539 |
+
"learning_rate": 0.00019143197311317014,
|
| 540 |
+
"loss": 1.1275,
|
| 541 |
+
"step": 375
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.22238783450760571,
|
| 545 |
+
"grad_norm": 0.06787109375,
|
| 546 |
+
"learning_rate": 0.00019101330719112705,
|
| 547 |
+
"loss": 1.1684,
|
| 548 |
+
"step": 380
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.22531399022481105,
|
| 552 |
+
"grad_norm": 0.07763671875,
|
| 553 |
+
"learning_rate": 0.00019058513537141428,
|
| 554 |
+
"loss": 1.1606,
|
| 555 |
+
"step": 385
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.2282401459420164,
|
| 559 |
+
"grad_norm": 0.07421875,
|
| 560 |
+
"learning_rate": 0.0001901475023744977,
|
| 561 |
+
"loss": 1.148,
|
| 562 |
+
"step": 390
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.23116630165922172,
|
| 566 |
+
"grad_norm": 0.0693359375,
|
| 567 |
+
"learning_rate": 0.00018970045390901728,
|
| 568 |
+
"loss": 1.1626,
|
| 569 |
+
"step": 395
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.23409245737642706,
|
| 573 |
+
"grad_norm": 0.0849609375,
|
| 574 |
+
"learning_rate": 0.00018924403666701286,
|
| 575 |
+
"loss": 1.1575,
|
| 576 |
+
"step": 400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.23701861309363242,
|
| 580 |
+
"grad_norm": 0.07470703125,
|
| 581 |
+
"learning_rate": 0.00018877829831904746,
|
| 582 |
+
"loss": 1.1637,
|
| 583 |
+
"step": 405
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.23994476881083776,
|
| 587 |
+
"grad_norm": 0.07568359375,
|
| 588 |
+
"learning_rate": 0.0001883032875092283,
|
| 589 |
+
"loss": 1.1441,
|
| 590 |
+
"step": 410
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.2428709245280431,
|
| 594 |
+
"grad_norm": 0.0771484375,
|
| 595 |
+
"learning_rate": 0.00018781905385012627,
|
| 596 |
+
"loss": 1.1615,
|
| 597 |
+
"step": 415
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.24579708024524843,
|
| 601 |
+
"grad_norm": 0.06884765625,
|
| 602 |
+
"learning_rate": 0.000187325647917594,
|
| 603 |
+
"loss": 1.1536,
|
| 604 |
+
"step": 420
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.24872323596245377,
|
| 608 |
+
"grad_norm": 0.08203125,
|
| 609 |
+
"learning_rate": 0.00018682312124548346,
|
| 610 |
+
"loss": 1.1512,
|
| 611 |
+
"step": 425
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.2516493916796591,
|
| 615 |
+
"grad_norm": 0.07470703125,
|
| 616 |
+
"learning_rate": 0.00018631152632026364,
|
| 617 |
+
"loss": 1.1397,
|
| 618 |
+
"step": 430
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.25457554739686444,
|
| 622 |
+
"grad_norm": 0.07958984375,
|
| 623 |
+
"learning_rate": 0.00018579091657553844,
|
| 624 |
+
"loss": 1.1585,
|
| 625 |
+
"step": 435
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.2575017031140698,
|
| 629 |
+
"grad_norm": 0.08154296875,
|
| 630 |
+
"learning_rate": 0.00018526134638646583,
|
| 631 |
+
"loss": 1.1612,
|
| 632 |
+
"step": 440
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.2604278588312751,
|
| 636 |
+
"grad_norm": 0.07470703125,
|
| 637 |
+
"learning_rate": 0.00018472287106407876,
|
| 638 |
+
"loss": 1.1272,
|
| 639 |
+
"step": 445
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.26335401454848045,
|
| 643 |
+
"grad_norm": 0.09375,
|
| 644 |
+
"learning_rate": 0.00018417554684950794,
|
| 645 |
+
"loss": 1.1413,
|
| 646 |
+
"step": 450
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.2662801702656858,
|
| 650 |
+
"grad_norm": 0.0703125,
|
| 651 |
+
"learning_rate": 0.00018361943090810796,
|
| 652 |
+
"loss": 1.1489,
|
| 653 |
+
"step": 455
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.2692063259828911,
|
| 657 |
+
"grad_norm": 0.0888671875,
|
| 658 |
+
"learning_rate": 0.00018305458132348657,
|
| 659 |
+
"loss": 1.1575,
|
| 660 |
+
"step": 460
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.27213248170009646,
|
| 664 |
+
"grad_norm": 0.07666015625,
|
| 665 |
+
"learning_rate": 0.00018248105709143799,
|
| 666 |
+
"loss": 1.136,
|
| 667 |
+
"step": 465
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.2750586374173018,
|
| 671 |
+
"grad_norm": 0.08203125,
|
| 672 |
+
"learning_rate": 0.00018189891811378137,
|
| 673 |
+
"loss": 1.1369,
|
| 674 |
+
"step": 470
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.27798479313450714,
|
| 678 |
+
"grad_norm": 0.08251953125,
|
| 679 |
+
"learning_rate": 0.0001813082251921041,
|
| 680 |
+
"loss": 1.1255,
|
| 681 |
+
"step": 475
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.28091094885171247,
|
| 685 |
+
"grad_norm": 0.09375,
|
| 686 |
+
"learning_rate": 0.0001807090400214114,
|
| 687 |
+
"loss": 1.1288,
|
| 688 |
+
"step": 480
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.2838371045689178,
|
| 692 |
+
"grad_norm": 0.07958984375,
|
| 693 |
+
"learning_rate": 0.00018010142518368278,
|
| 694 |
+
"loss": 1.1233,
|
| 695 |
+
"step": 485
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.28676326028612315,
|
| 699 |
+
"grad_norm": 0.078125,
|
| 700 |
+
"learning_rate": 0.00017948544414133534,
|
| 701 |
+
"loss": 1.1475,
|
| 702 |
+
"step": 490
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.2896894160033285,
|
| 706 |
+
"grad_norm": 0.080078125,
|
| 707 |
+
"learning_rate": 0.00017886116123059574,
|
| 708 |
+
"loss": 1.1356,
|
| 709 |
+
"step": 495
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.2926155717205338,
|
| 713 |
+
"grad_norm": 0.07568359375,
|
| 714 |
+
"learning_rate": 0.00017822864165478034,
|
| 715 |
+
"loss": 1.1553,
|
| 716 |
+
"step": 500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.29554172743773915,
|
| 720 |
+
"grad_norm": 0.08154296875,
|
| 721 |
+
"learning_rate": 0.00017758795147748523,
|
| 722 |
+
"loss": 1.1188,
|
| 723 |
+
"step": 505
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.2984678831549445,
|
| 727 |
+
"grad_norm": 0.07275390625,
|
| 728 |
+
"learning_rate": 0.00017693915761568608,
|
| 729 |
+
"loss": 1.1388,
|
| 730 |
+
"step": 510
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.3013940388721499,
|
| 734 |
+
"grad_norm": 0.07568359375,
|
| 735 |
+
"learning_rate": 0.000176282327832749,
|
| 736 |
+
"loss": 1.1267,
|
| 737 |
+
"step": 515
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.3043201945893552,
|
| 741 |
+
"grad_norm": 0.083984375,
|
| 742 |
+
"learning_rate": 0.0001756175307313531,
|
| 743 |
+
"loss": 1.1341,
|
| 744 |
+
"step": 520
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.30724635030656056,
|
| 748 |
+
"grad_norm": 0.08349609375,
|
| 749 |
+
"learning_rate": 0.00017494483574632513,
|
| 750 |
+
"loss": 1.1365,
|
| 751 |
+
"step": 525
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.3101725060237659,
|
| 755 |
+
"grad_norm": 0.08447265625,
|
| 756 |
+
"learning_rate": 0.00017426431313738734,
|
| 757 |
+
"loss": 1.1335,
|
| 758 |
+
"step": 530
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.31309866174097123,
|
| 762 |
+
"grad_norm": 0.0703125,
|
| 763 |
+
"learning_rate": 0.00017357603398181936,
|
| 764 |
+
"loss": 1.1484,
|
| 765 |
+
"step": 535
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.31602481745817657,
|
| 769 |
+
"grad_norm": 0.07568359375,
|
| 770 |
+
"learning_rate": 0.00017288007016703444,
|
| 771 |
+
"loss": 1.1186,
|
| 772 |
+
"step": 540
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.3189509731753819,
|
| 776 |
+
"grad_norm": 0.06982421875,
|
| 777 |
+
"learning_rate": 0.00017217649438307106,
|
| 778 |
+
"loss": 1.1442,
|
| 779 |
+
"step": 545
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.32187712889258724,
|
| 783 |
+
"grad_norm": 0.07080078125,
|
| 784 |
+
"learning_rate": 0.00017146538011500093,
|
| 785 |
+
"loss": 1.1284,
|
| 786 |
+
"step": 550
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.3248032846097926,
|
| 790 |
+
"grad_norm": 0.0693359375,
|
| 791 |
+
"learning_rate": 0.00017074680163525375,
|
| 792 |
+
"loss": 1.1331,
|
| 793 |
+
"step": 555
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.3277294403269979,
|
| 797 |
+
"grad_norm": 0.07421875,
|
| 798 |
+
"learning_rate": 0.00017002083399586,
|
| 799 |
+
"loss": 1.1255,
|
| 800 |
+
"step": 560
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.33065559604420325,
|
| 804 |
+
"grad_norm": 0.07421875,
|
| 805 |
+
"learning_rate": 0.00016928755302061173,
|
| 806 |
+
"loss": 1.1354,
|
| 807 |
+
"step": 565
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.3335817517614086,
|
| 811 |
+
"grad_norm": 0.078125,
|
| 812 |
+
"learning_rate": 0.0001685470352971437,
|
| 813 |
+
"loss": 1.1333,
|
| 814 |
+
"step": 570
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.3365079074786139,
|
| 818 |
+
"grad_norm": 0.07470703125,
|
| 819 |
+
"learning_rate": 0.00016779935816893353,
|
| 820 |
+
"loss": 1.1376,
|
| 821 |
+
"step": 575
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.33943406319581926,
|
| 825 |
+
"grad_norm": 0.06982421875,
|
| 826 |
+
"learning_rate": 0.00016704459972722414,
|
| 827 |
+
"loss": 1.1249,
|
| 828 |
+
"step": 580
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.3423602189130246,
|
| 832 |
+
"grad_norm": 0.0732421875,
|
| 833 |
+
"learning_rate": 0.00016628283880286703,
|
| 834 |
+
"loss": 1.1451,
|
| 835 |
+
"step": 585
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.34528637463022993,
|
| 839 |
+
"grad_norm": 0.0693359375,
|
| 840 |
+
"learning_rate": 0.00016551415495808915,
|
| 841 |
+
"loss": 1.1195,
|
| 842 |
+
"step": 590
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.34821253034743527,
|
| 846 |
+
"grad_norm": 0.0810546875,
|
| 847 |
+
"learning_rate": 0.00016473862847818277,
|
| 848 |
+
"loss": 1.146,
|
| 849 |
+
"step": 595
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.3511386860646406,
|
| 853 |
+
"grad_norm": 0.07470703125,
|
| 854 |
+
"learning_rate": 0.00016395634036312013,
|
| 855 |
+
"loss": 1.1327,
|
| 856 |
+
"step": 600
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.35406484178184594,
|
| 860 |
+
"grad_norm": 0.0791015625,
|
| 861 |
+
"learning_rate": 0.00016316737231909342,
|
| 862 |
+
"loss": 1.1176,
|
| 863 |
+
"step": 605
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.3569909974990513,
|
| 867 |
+
"grad_norm": 0.07470703125,
|
| 868 |
+
"learning_rate": 0.000162371806749981,
|
| 869 |
+
"loss": 1.1208,
|
| 870 |
+
"step": 610
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.3599171532162566,
|
| 874 |
+
"grad_norm": 0.07568359375,
|
| 875 |
+
"learning_rate": 0.00016156972674874056,
|
| 876 |
+
"loss": 1.1315,
|
| 877 |
+
"step": 615
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.36284330893346195,
|
| 881 |
+
"grad_norm": 0.06982421875,
|
| 882 |
+
"learning_rate": 0.00016076121608873072,
|
| 883 |
+
"loss": 1.1455,
|
| 884 |
+
"step": 620
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.3657694646506673,
|
| 888 |
+
"grad_norm": 0.072265625,
|
| 889 |
+
"learning_rate": 0.000159946359214961,
|
| 890 |
+
"loss": 1.1234,
|
| 891 |
+
"step": 625
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.3686956203678726,
|
| 895 |
+
"grad_norm": 0.07421875,
|
| 896 |
+
"learning_rate": 0.00015912524123527221,
|
| 897 |
+
"loss": 1.1185,
|
| 898 |
+
"step": 630
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.37162177608507796,
|
| 902 |
+
"grad_norm": 0.0693359375,
|
| 903 |
+
"learning_rate": 0.0001582979479114472,
|
| 904 |
+
"loss": 1.1208,
|
| 905 |
+
"step": 635
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.3745479318022833,
|
| 909 |
+
"grad_norm": 0.07666015625,
|
| 910 |
+
"learning_rate": 0.0001574645656502536,
|
| 911 |
+
"loss": 1.1257,
|
| 912 |
+
"step": 640
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.37747408751948863,
|
| 916 |
+
"grad_norm": 0.07421875,
|
| 917 |
+
"learning_rate": 0.0001566251814944188,
|
| 918 |
+
"loss": 1.1317,
|
| 919 |
+
"step": 645
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.38040024323669397,
|
| 923 |
+
"grad_norm": 0.07373046875,
|
| 924 |
+
"learning_rate": 0.00015577988311353904,
|
| 925 |
+
"loss": 1.1431,
|
| 926 |
+
"step": 650
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.3833263989538993,
|
| 930 |
+
"grad_norm": 0.0771484375,
|
| 931 |
+
"learning_rate": 0.0001549287587949226,
|
| 932 |
+
"loss": 1.1253,
|
| 933 |
+
"step": 655
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.38625255467110464,
|
| 937 |
+
"grad_norm": 0.07080078125,
|
| 938 |
+
"learning_rate": 0.00015407189743436864,
|
| 939 |
+
"loss": 1.1314,
|
| 940 |
+
"step": 660
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.38917871038831003,
|
| 944 |
+
"grad_norm": 0.07275390625,
|
| 945 |
+
"learning_rate": 0.00015320938852688248,
|
| 946 |
+
"loss": 1.1148,
|
| 947 |
+
"step": 665
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.39210486610551537,
|
| 951 |
+
"grad_norm": 0.07080078125,
|
| 952 |
+
"learning_rate": 0.00015234132215732822,
|
| 953 |
+
"loss": 1.141,
|
| 954 |
+
"step": 670
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.3950310218227207,
|
| 958 |
+
"grad_norm": 0.072265625,
|
| 959 |
+
"learning_rate": 0.00015146778899102,
|
| 960 |
+
"loss": 1.1222,
|
| 961 |
+
"step": 675
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.39795717753992604,
|
| 965 |
+
"grad_norm": 0.07568359375,
|
| 966 |
+
"learning_rate": 0.00015058888026425212,
|
| 967 |
+
"loss": 1.1177,
|
| 968 |
+
"step": 680
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.4008833332571314,
|
| 972 |
+
"grad_norm": 0.0830078125,
|
| 973 |
+
"learning_rate": 0.00014970468777477026,
|
| 974 |
+
"loss": 1.1181,
|
| 975 |
+
"step": 685
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.4038094889743367,
|
| 979 |
+
"grad_norm": 0.08349609375,
|
| 980 |
+
"learning_rate": 0.00014881530387218325,
|
| 981 |
+
"loss": 1.1417,
|
| 982 |
+
"step": 690
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.40673564469154205,
|
| 986 |
+
"grad_norm": 0.08154296875,
|
| 987 |
+
"learning_rate": 0.00014792082144831793,
|
| 988 |
+
"loss": 1.1302,
|
| 989 |
+
"step": 695
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.4096618004087474,
|
| 993 |
+
"grad_norm": 0.06787109375,
|
| 994 |
+
"learning_rate": 0.00014702133392751688,
|
| 995 |
+
"loss": 1.122,
|
| 996 |
+
"step": 700
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.4125879561259527,
|
| 1000 |
+
"grad_norm": 0.07177734375,
|
| 1001 |
+
"learning_rate": 0.00014611693525688066,
|
| 1002 |
+
"loss": 1.1268,
|
| 1003 |
+
"step": 705
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.41551411184315806,
|
| 1007 |
+
"grad_norm": 0.06982421875,
|
| 1008 |
+
"learning_rate": 0.00014520771989645563,
|
| 1009 |
+
"loss": 1.1238,
|
| 1010 |
+
"step": 710
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.4184402675603634,
|
| 1014 |
+
"grad_norm": 0.07177734375,
|
| 1015 |
+
"learning_rate": 0.00014429378280936804,
|
| 1016 |
+
"loss": 1.119,
|
| 1017 |
+
"step": 715
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.42136642327756874,
|
| 1021 |
+
"grad_norm": 0.0703125,
|
| 1022 |
+
"learning_rate": 0.0001433752194519054,
|
| 1023 |
+
"loss": 1.1187,
|
| 1024 |
+
"step": 720
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.42429257899477407,
|
| 1028 |
+
"grad_norm": 0.0712890625,
|
| 1029 |
+
"learning_rate": 0.00014245212576354682,
|
| 1030 |
+
"loss": 1.122,
|
| 1031 |
+
"step": 725
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.4272187347119794,
|
| 1035 |
+
"grad_norm": 0.0751953125,
|
| 1036 |
+
"learning_rate": 0.0001415245981569424,
|
| 1037 |
+
"loss": 1.1267,
|
| 1038 |
+
"step": 730
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.43014489042918475,
|
| 1042 |
+
"grad_norm": 0.06591796875,
|
| 1043 |
+
"learning_rate": 0.00014059273350784342,
|
| 1044 |
+
"loss": 1.1273,
|
| 1045 |
+
"step": 735
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.4330710461463901,
|
| 1049 |
+
"grad_norm": 0.076171875,
|
| 1050 |
+
"learning_rate": 0.00013965662914498428,
|
| 1051 |
+
"loss": 1.1267,
|
| 1052 |
+
"step": 740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.4359972018635954,
|
| 1056 |
+
"grad_norm": 0.06884765625,
|
| 1057 |
+
"learning_rate": 0.00013871638283991677,
|
| 1058 |
+
"loss": 1.1175,
|
| 1059 |
+
"step": 745
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.43892335758080075,
|
| 1063 |
+
"grad_norm": 0.06787109375,
|
| 1064 |
+
"learning_rate": 0.0001377720927967985,
|
| 1065 |
+
"loss": 1.1211,
|
| 1066 |
+
"step": 750
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.4418495132980061,
|
| 1070 |
+
"grad_norm": 0.1884765625,
|
| 1071 |
+
"learning_rate": 0.00013682385764213572,
|
| 1072 |
+
"loss": 1.1319,
|
| 1073 |
+
"step": 755
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.44477566901521143,
|
| 1077 |
+
"grad_norm": 0.07666015625,
|
| 1078 |
+
"learning_rate": 0.00013587177641448265,
|
| 1079 |
+
"loss": 1.1233,
|
| 1080 |
+
"step": 760
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.44770182473241676,
|
| 1084 |
+
"grad_norm": 0.078125,
|
| 1085 |
+
"learning_rate": 0.00013491594855409697,
|
| 1086 |
+
"loss": 1.1385,
|
| 1087 |
+
"step": 765
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.4506279804496221,
|
| 1091 |
+
"grad_norm": 0.07470703125,
|
| 1092 |
+
"learning_rate": 0.00013395647389255396,
|
| 1093 |
+
"loss": 1.1189,
|
| 1094 |
+
"step": 770
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.45355413616682744,
|
| 1098 |
+
"grad_norm": 0.072265625,
|
| 1099 |
+
"learning_rate": 0.00013299345264231957,
|
| 1100 |
+
"loss": 1.1157,
|
| 1101 |
+
"step": 775
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.4564802918840328,
|
| 1105 |
+
"grad_norm": 0.07275390625,
|
| 1106 |
+
"learning_rate": 0.00013202698538628376,
|
| 1107 |
+
"loss": 1.1224,
|
| 1108 |
+
"step": 780
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.4594064476012381,
|
| 1112 |
+
"grad_norm": 0.07080078125,
|
| 1113 |
+
"learning_rate": 0.00013105717306725501,
|
| 1114 |
+
"loss": 1.1283,
|
| 1115 |
+
"step": 785
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.46233260331844345,
|
| 1119 |
+
"grad_norm": 0.0703125,
|
| 1120 |
+
"learning_rate": 0.0001300841169774174,
|
| 1121 |
+
"loss": 1.131,
|
| 1122 |
+
"step": 790
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.4652587590356488,
|
| 1126 |
+
"grad_norm": 0.0712890625,
|
| 1127 |
+
"learning_rate": 0.000129107918747751,
|
| 1128 |
+
"loss": 1.1175,
|
| 1129 |
+
"step": 795
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.4681849147528541,
|
| 1133 |
+
"grad_norm": 0.07373046875,
|
| 1134 |
+
"learning_rate": 0.00012812868033741724,
|
| 1135 |
+
"loss": 1.138,
|
| 1136 |
+
"step": 800
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.47111107047005946,
|
| 1140 |
+
"grad_norm": 0.07275390625,
|
| 1141 |
+
"learning_rate": 0.00012714650402310967,
|
| 1142 |
+
"loss": 1.1344,
|
| 1143 |
+
"step": 805
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.47403722618726485,
|
| 1147 |
+
"grad_norm": 0.07470703125,
|
| 1148 |
+
"learning_rate": 0.00012616149238837146,
|
| 1149 |
+
"loss": 1.1195,
|
| 1150 |
+
"step": 810
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.4769633819044702,
|
| 1154 |
+
"grad_norm": 0.07958984375,
|
| 1155 |
+
"learning_rate": 0.00012517374831288146,
|
| 1156 |
+
"loss": 1.1005,
|
| 1157 |
+
"step": 815
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.4798895376216755,
|
| 1161 |
+
"grad_norm": 0.0712890625,
|
| 1162 |
+
"learning_rate": 0.00012418337496170842,
|
| 1163 |
+
"loss": 1.1158,
|
| 1164 |
+
"step": 820
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.48281569333888086,
|
| 1168 |
+
"grad_norm": 0.0712890625,
|
| 1169 |
+
"learning_rate": 0.00012319047577453638,
|
| 1170 |
+
"loss": 1.1181,
|
| 1171 |
+
"step": 825
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.4857418490560862,
|
| 1175 |
+
"grad_norm": 0.07177734375,
|
| 1176 |
+
"learning_rate": 0.00012219515445486054,
|
| 1177 |
+
"loss": 1.1321,
|
| 1178 |
+
"step": 830
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.48866800477329153,
|
| 1182 |
+
"grad_norm": 0.0751953125,
|
| 1183 |
+
"learning_rate": 0.00012119751495915617,
|
| 1184 |
+
"loss": 1.1309,
|
| 1185 |
+
"step": 835
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.49159416049049687,
|
| 1189 |
+
"grad_norm": 0.06884765625,
|
| 1190 |
+
"learning_rate": 0.00012019766148602062,
|
| 1191 |
+
"loss": 1.1276,
|
| 1192 |
+
"step": 840
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.4945203162077022,
|
| 1196 |
+
"grad_norm": 0.068359375,
|
| 1197 |
+
"learning_rate": 0.00011919569846529057,
|
| 1198 |
+
"loss": 1.1173,
|
| 1199 |
+
"step": 845
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.49744647192490754,
|
| 1203 |
+
"grad_norm": 0.0703125,
|
| 1204 |
+
"learning_rate": 0.00011819173054713466,
|
| 1205 |
+
"loss": 1.1111,
|
| 1206 |
+
"step": 850
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.5003726276421129,
|
| 1210 |
+
"grad_norm": 0.0751953125,
|
| 1211 |
+
"learning_rate": 0.00011718586259112326,
|
| 1212 |
+
"loss": 1.1137,
|
| 1213 |
+
"step": 855
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.5032987833593182,
|
| 1217 |
+
"grad_norm": 0.0751953125,
|
| 1218 |
+
"learning_rate": 0.0001161781996552765,
|
| 1219 |
+
"loss": 1.1157,
|
| 1220 |
+
"step": 860
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.5062249390765236,
|
| 1224 |
+
"grad_norm": 0.0703125,
|
| 1225 |
+
"learning_rate": 0.00011516884698509143,
|
| 1226 |
+
"loss": 1.1136,
|
| 1227 |
+
"step": 865
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.5091510947937289,
|
| 1231 |
+
"grad_norm": 0.07568359375,
|
| 1232 |
+
"learning_rate": 0.00011415791000254964,
|
| 1233 |
+
"loss": 1.1217,
|
| 1234 |
+
"step": 870
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.5120772505109342,
|
| 1238 |
+
"grad_norm": 0.07275390625,
|
| 1239 |
+
"learning_rate": 0.0001131454942951065,
|
| 1240 |
+
"loss": 1.119,
|
| 1241 |
+
"step": 875
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.5150034062281396,
|
| 1245 |
+
"grad_norm": 0.0693359375,
|
| 1246 |
+
"learning_rate": 0.0001121317056046629,
|
| 1247 |
+
"loss": 1.1122,
|
| 1248 |
+
"step": 880
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.5179295619453449,
|
| 1252 |
+
"grad_norm": 0.072265625,
|
| 1253 |
+
"learning_rate": 0.00011111664981652121,
|
| 1254 |
+
"loss": 1.1137,
|
| 1255 |
+
"step": 885
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.5208557176625502,
|
| 1259 |
+
"grad_norm": 0.0693359375,
|
| 1260 |
+
"learning_rate": 0.00011010043294832601,
|
| 1261 |
+
"loss": 1.1132,
|
| 1262 |
+
"step": 890
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.5237818733797556,
|
| 1266 |
+
"grad_norm": 0.072265625,
|
| 1267 |
+
"learning_rate": 0.00010908316113899097,
|
| 1268 |
+
"loss": 1.1373,
|
| 1269 |
+
"step": 895
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.5267080290969609,
|
| 1273 |
+
"grad_norm": 0.0732421875,
|
| 1274 |
+
"learning_rate": 0.00010806494063761335,
|
| 1275 |
+
"loss": 1.1165,
|
| 1276 |
+
"step": 900
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.5296341848141662,
|
| 1280 |
+
"grad_norm": 0.0703125,
|
| 1281 |
+
"learning_rate": 0.00010704587779237654,
|
| 1282 |
+
"loss": 1.1149,
|
| 1283 |
+
"step": 905
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.5325603405313716,
|
| 1287 |
+
"grad_norm": 0.06982421875,
|
| 1288 |
+
"learning_rate": 0.00010602607903944279,
|
| 1289 |
+
"loss": 1.1244,
|
| 1290 |
+
"step": 910
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.5354864962485769,
|
| 1294 |
+
"grad_norm": 0.06787109375,
|
| 1295 |
+
"learning_rate": 0.00010500565089183627,
|
| 1296 |
+
"loss": 1.1141,
|
| 1297 |
+
"step": 915
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.5384126519657823,
|
| 1301 |
+
"grad_norm": 0.0673828125,
|
| 1302 |
+
"learning_rate": 0.00010398469992831832,
|
| 1303 |
+
"loss": 1.1031,
|
| 1304 |
+
"step": 920
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.5413388076829876,
|
| 1308 |
+
"grad_norm": 0.06787109375,
|
| 1309 |
+
"learning_rate": 0.00010296333278225599,
|
| 1310 |
+
"loss": 1.1072,
|
| 1311 |
+
"step": 925
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.5442649634001929,
|
| 1315 |
+
"grad_norm": 0.06982421875,
|
| 1316 |
+
"learning_rate": 0.00010194165613048444,
|
| 1317 |
+
"loss": 1.0993,
|
| 1318 |
+
"step": 930
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.5471911191173983,
|
| 1322 |
+
"grad_norm": 0.07275390625,
|
| 1323 |
+
"learning_rate": 0.00010091977668216524,
|
| 1324 |
+
"loss": 1.1089,
|
| 1325 |
+
"step": 935
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.5501172748346036,
|
| 1329 |
+
"grad_norm": 0.068359375,
|
| 1330 |
+
"learning_rate": 9.989780116764115e-05,
|
| 1331 |
+
"loss": 1.1042,
|
| 1332 |
+
"step": 940
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.5530434305518089,
|
| 1336 |
+
"grad_norm": 0.0712890625,
|
| 1337 |
+
"learning_rate": 9.887583632728845e-05,
|
| 1338 |
+
"loss": 1.1062,
|
| 1339 |
+
"step": 945
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.5559695862690143,
|
| 1343 |
+
"grad_norm": 0.07080078125,
|
| 1344 |
+
"learning_rate": 9.785398890036867e-05,
|
| 1345 |
+
"loss": 1.1092,
|
| 1346 |
+
"step": 950
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.5588957419862196,
|
| 1350 |
+
"grad_norm": 0.0693359375,
|
| 1351 |
+
"learning_rate": 9.683236561388e-05,
|
| 1352 |
+
"loss": 1.1173,
|
| 1353 |
+
"step": 955
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.5618218977034249,
|
| 1357 |
+
"grad_norm": 0.0703125,
|
| 1358 |
+
"learning_rate": 9.581107317141026e-05,
|
| 1359 |
+
"loss": 1.1407,
|
| 1360 |
+
"step": 960
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.5647480534206303,
|
| 1364 |
+
"grad_norm": 0.07666015625,
|
| 1365 |
+
"learning_rate": 9.479021824199229e-05,
|
| 1366 |
+
"loss": 1.1365,
|
| 1367 |
+
"step": 965
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.5676742091378356,
|
| 1371 |
+
"grad_norm": 0.07763671875,
|
| 1372 |
+
"learning_rate": 9.376990744896276e-05,
|
| 1373 |
+
"loss": 1.1031,
|
| 1374 |
+
"step": 970
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.570600364855041,
|
| 1378 |
+
"grad_norm": 0.0703125,
|
| 1379 |
+
"learning_rate": 9.275024735882588e-05,
|
| 1380 |
+
"loss": 1.0896,
|
| 1381 |
+
"step": 975
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.5735265205722463,
|
| 1385 |
+
"grad_norm": 0.07275390625,
|
| 1386 |
+
"learning_rate": 9.173134447012322e-05,
|
| 1387 |
+
"loss": 1.1094,
|
| 1388 |
+
"step": 980
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.5764526762894516,
|
| 1392 |
+
"grad_norm": 0.0751953125,
|
| 1393 |
+
"learning_rate": 9.071330520231033e-05,
|
| 1394 |
+
"loss": 1.1127,
|
| 1395 |
+
"step": 985
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 0.579378832006657,
|
| 1399 |
+
"grad_norm": 0.0693359375,
|
| 1400 |
+
"learning_rate": 8.969623588464163e-05,
|
| 1401 |
+
"loss": 1.1176,
|
| 1402 |
+
"step": 990
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 0.5823049877238623,
|
| 1406 |
+
"grad_norm": 0.0703125,
|
| 1407 |
+
"learning_rate": 8.868024274506505e-05,
|
| 1408 |
+
"loss": 1.112,
|
| 1409 |
+
"step": 995
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 0.5852311434410676,
|
| 1413 |
+
"grad_norm": 0.0712890625,
|
| 1414 |
+
"learning_rate": 8.766543189912705e-05,
|
| 1415 |
+
"loss": 1.0846,
|
| 1416 |
+
"step": 1000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.588157299158273,
|
| 1420 |
+
"grad_norm": 0.07177734375,
|
| 1421 |
+
"learning_rate": 8.665190933888904e-05,
|
| 1422 |
+
"loss": 1.0961,
|
| 1423 |
+
"step": 1005
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 0.5910834548754783,
|
| 1427 |
+
"grad_norm": 0.0712890625,
|
| 1428 |
+
"learning_rate": 8.56397809218574e-05,
|
| 1429 |
+
"loss": 1.1146,
|
| 1430 |
+
"step": 1010
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.5940096105926836,
|
| 1434 |
+
"grad_norm": 0.07568359375,
|
| 1435 |
+
"learning_rate": 8.4629152359927e-05,
|
| 1436 |
+
"loss": 1.1066,
|
| 1437 |
+
"step": 1015
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 0.596935766309889,
|
| 1441 |
+
"grad_norm": 0.07080078125,
|
| 1442 |
+
"learning_rate": 8.362012920834014e-05,
|
| 1443 |
+
"loss": 1.1253,
|
| 1444 |
+
"step": 1020
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 0.5998619220270943,
|
| 1448 |
+
"grad_norm": 0.0693359375,
|
| 1449 |
+
"learning_rate": 8.261281685466177e-05,
|
| 1450 |
+
"loss": 1.1072,
|
| 1451 |
+
"step": 1025
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 0.6027880777442998,
|
| 1455 |
+
"grad_norm": 0.07373046875,
|
| 1456 |
+
"learning_rate": 8.160732050777235e-05,
|
| 1457 |
+
"loss": 1.1147,
|
| 1458 |
+
"step": 1030
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.6057142334615051,
|
| 1462 |
+
"grad_norm": 0.0712890625,
|
| 1463 |
+
"learning_rate": 8.060374518687926e-05,
|
| 1464 |
+
"loss": 1.11,
|
| 1465 |
+
"step": 1035
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.6086403891787104,
|
| 1469 |
+
"grad_norm": 0.07080078125,
|
| 1470 |
+
"learning_rate": 7.960219571054799e-05,
|
| 1471 |
+
"loss": 1.123,
|
| 1472 |
+
"step": 1040
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.6115665448959158,
|
| 1476 |
+
"grad_norm": 0.06787109375,
|
| 1477 |
+
"learning_rate": 7.860277668575449e-05,
|
| 1478 |
+
"loss": 1.1035,
|
| 1479 |
+
"step": 1045
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 0.6144927006131211,
|
| 1483 |
+
"grad_norm": 0.0712890625,
|
| 1484 |
+
"learning_rate": 7.76055924969594e-05,
|
| 1485 |
+
"loss": 1.09,
|
| 1486 |
+
"step": 1050
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 0.6174188563303264,
|
| 1490 |
+
"grad_norm": 0.0712890625,
|
| 1491 |
+
"learning_rate": 7.661074729520548e-05,
|
| 1492 |
+
"loss": 1.1279,
|
| 1493 |
+
"step": 1055
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.6203450120475318,
|
| 1497 |
+
"grad_norm": 0.072265625,
|
| 1498 |
+
"learning_rate": 7.561834498723974e-05,
|
| 1499 |
+
"loss": 1.1141,
|
| 1500 |
+
"step": 1060
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 0.6232711677647371,
|
| 1504 |
+
"grad_norm": 0.068359375,
|
| 1505 |
+
"learning_rate": 7.462848922466092e-05,
|
| 1506 |
+
"loss": 1.1102,
|
| 1507 |
+
"step": 1065
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 0.6261973234819425,
|
| 1511 |
+
"grad_norm": 0.0673828125,
|
| 1512 |
+
"learning_rate": 7.364128339309326e-05,
|
| 1513 |
+
"loss": 1.1128,
|
| 1514 |
+
"step": 1070
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 0.6291234791991478,
|
| 1518 |
+
"grad_norm": 0.07275390625,
|
| 1519 |
+
"learning_rate": 7.265683060138868e-05,
|
| 1520 |
+
"loss": 1.1054,
|
| 1521 |
+
"step": 1075
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 0.6320496349163531,
|
| 1525 |
+
"grad_norm": 0.07666015625,
|
| 1526 |
+
"learning_rate": 7.167523367085749e-05,
|
| 1527 |
+
"loss": 1.1097,
|
| 1528 |
+
"step": 1080
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 0.6349757906335585,
|
| 1532 |
+
"grad_norm": 0.0673828125,
|
| 1533 |
+
"learning_rate": 7.069659512452918e-05,
|
| 1534 |
+
"loss": 1.1148,
|
| 1535 |
+
"step": 1085
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 0.6379019463507638,
|
| 1539 |
+
"grad_norm": 0.06982421875,
|
| 1540 |
+
"learning_rate": 6.972101717644429e-05,
|
| 1541 |
+
"loss": 1.0997,
|
| 1542 |
+
"step": 1090
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 0.6408281020679691,
|
| 1546 |
+
"grad_norm": 0.07080078125,
|
| 1547 |
+
"learning_rate": 6.874860172097883e-05,
|
| 1548 |
+
"loss": 1.097,
|
| 1549 |
+
"step": 1095
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.6437542577851745,
|
| 1553 |
+
"grad_norm": 0.07177734375,
|
| 1554 |
+
"learning_rate": 6.777945032220187e-05,
|
| 1555 |
+
"loss": 1.1006,
|
| 1556 |
+
"step": 1100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 0.6466804135023798,
|
| 1560 |
+
"grad_norm": 0.0693359375,
|
| 1561 |
+
"learning_rate": 6.681366420326747e-05,
|
| 1562 |
+
"loss": 1.1191,
|
| 1563 |
+
"step": 1105
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 0.6496065692195852,
|
| 1567 |
+
"grad_norm": 0.162109375,
|
| 1568 |
+
"learning_rate": 6.58513442358427e-05,
|
| 1569 |
+
"loss": 1.0901,
|
| 1570 |
+
"step": 1110
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 0.6525327249367905,
|
| 1574 |
+
"grad_norm": 0.0712890625,
|
| 1575 |
+
"learning_rate": 6.489259092957193e-05,
|
| 1576 |
+
"loss": 1.1113,
|
| 1577 |
+
"step": 1115
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.6554588806539958,
|
| 1581 |
+
"grad_norm": 0.0751953125,
|
| 1582 |
+
"learning_rate": 6.3937504421579e-05,
|
| 1583 |
+
"loss": 1.0945,
|
| 1584 |
+
"step": 1120
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.6583850363712012,
|
| 1588 |
+
"grad_norm": 0.06884765625,
|
| 1589 |
+
"learning_rate": 6.298618446600856e-05,
|
| 1590 |
+
"loss": 1.1073,
|
| 1591 |
+
"step": 1125
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 0.6613111920884065,
|
| 1595 |
+
"grad_norm": 0.0693359375,
|
| 1596 |
+
"learning_rate": 6.203873042360722e-05,
|
| 1597 |
+
"loss": 1.1178,
|
| 1598 |
+
"step": 1130
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 0.6642373478056118,
|
| 1602 |
+
"grad_norm": 0.072265625,
|
| 1603 |
+
"learning_rate": 6.109524125134571e-05,
|
| 1604 |
+
"loss": 1.1291,
|
| 1605 |
+
"step": 1135
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 0.6671635035228172,
|
| 1609 |
+
"grad_norm": 0.07080078125,
|
| 1610 |
+
"learning_rate": 6.015581549208322e-05,
|
| 1611 |
+
"loss": 1.0985,
|
| 1612 |
+
"step": 1140
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 0.6700896592400225,
|
| 1616 |
+
"grad_norm": 0.06884765625,
|
| 1617 |
+
"learning_rate": 5.9220551264275356e-05,
|
| 1618 |
+
"loss": 1.1294,
|
| 1619 |
+
"step": 1145
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 0.6730158149572278,
|
| 1623 |
+
"grad_norm": 0.06884765625,
|
| 1624 |
+
"learning_rate": 5.828954625172597e-05,
|
| 1625 |
+
"loss": 1.115,
|
| 1626 |
+
"step": 1150
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 0.6759419706744332,
|
| 1630 |
+
"grad_norm": 0.06689453125,
|
| 1631 |
+
"learning_rate": 5.736289769338441e-05,
|
| 1632 |
+
"loss": 1.1024,
|
| 1633 |
+
"step": 1155
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 0.6788681263916385,
|
| 1637 |
+
"grad_norm": 0.0703125,
|
| 1638 |
+
"learning_rate": 5.644070237318977e-05,
|
| 1639 |
+
"loss": 1.0993,
|
| 1640 |
+
"step": 1160
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 0.6817942821088439,
|
| 1644 |
+
"grad_norm": 0.0693359375,
|
| 1645 |
+
"learning_rate": 5.552305660996202e-05,
|
| 1646 |
+
"loss": 1.1172,
|
| 1647 |
+
"step": 1165
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 0.6847204378260492,
|
| 1651 |
+
"grad_norm": 0.0693359375,
|
| 1652 |
+
"learning_rate": 5.4610056247341814e-05,
|
| 1653 |
+
"loss": 1.0988,
|
| 1654 |
+
"step": 1170
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 0.6876465935432545,
|
| 1658 |
+
"grad_norm": 0.0712890625,
|
| 1659 |
+
"learning_rate": 5.3701796643780524e-05,
|
| 1660 |
+
"loss": 1.1142,
|
| 1661 |
+
"step": 1175
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 0.6905727492604599,
|
| 1665 |
+
"grad_norm": 0.0693359375,
|
| 1666 |
+
"learning_rate": 5.279837266258016e-05,
|
| 1667 |
+
"loss": 1.1271,
|
| 1668 |
+
"step": 1180
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 0.6934989049776652,
|
| 1672 |
+
"grad_norm": 0.06884765625,
|
| 1673 |
+
"learning_rate": 5.189987866198548e-05,
|
| 1674 |
+
"loss": 1.1055,
|
| 1675 |
+
"step": 1185
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 0.6964250606948705,
|
| 1679 |
+
"grad_norm": 0.06982421875,
|
| 1680 |
+
"learning_rate": 5.100640848532878e-05,
|
| 1681 |
+
"loss": 1.1277,
|
| 1682 |
+
"step": 1190
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 0.6993512164120759,
|
| 1686 |
+
"grad_norm": 0.06787109375,
|
| 1687 |
+
"learning_rate": 5.011805545122826e-05,
|
| 1688 |
+
"loss": 1.1093,
|
| 1689 |
+
"step": 1195
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.7022773721292812,
|
| 1693 |
+
"grad_norm": 0.06884765625,
|
| 1694 |
+
"learning_rate": 4.923491234384158e-05,
|
| 1695 |
+
"loss": 1.1055,
|
| 1696 |
+
"step": 1200
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.7052035278464865,
|
| 1700 |
+
"grad_norm": 0.0693359375,
|
| 1701 |
+
"learning_rate": 4.8357071403174746e-05,
|
| 1702 |
+
"loss": 1.098,
|
| 1703 |
+
"step": 1205
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 0.7081296835636919,
|
| 1707 |
+
"grad_norm": 0.06689453125,
|
| 1708 |
+
"learning_rate": 4.748462431544826e-05,
|
| 1709 |
+
"loss": 1.114,
|
| 1710 |
+
"step": 1210
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 0.7110558392808972,
|
| 1714 |
+
"grad_norm": 0.06884765625,
|
| 1715 |
+
"learning_rate": 4.661766220352097e-05,
|
| 1716 |
+
"loss": 1.1073,
|
| 1717 |
+
"step": 1215
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 0.7139819949981026,
|
| 1721 |
+
"grad_norm": 0.06982421875,
|
| 1722 |
+
"learning_rate": 4.5756275617372465e-05,
|
| 1723 |
+
"loss": 1.1121,
|
| 1724 |
+
"step": 1220
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 0.7169081507153079,
|
| 1728 |
+
"grad_norm": 0.0693359375,
|
| 1729 |
+
"learning_rate": 4.490055452464594e-05,
|
| 1730 |
+
"loss": 1.12,
|
| 1731 |
+
"step": 1225
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 0.7198343064325132,
|
| 1735 |
+
"grad_norm": 0.06787109375,
|
| 1736 |
+
"learning_rate": 4.405058830125137e-05,
|
| 1737 |
+
"loss": 1.092,
|
| 1738 |
+
"step": 1230
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 0.7227604621497186,
|
| 1742 |
+
"grad_norm": 0.06640625,
|
| 1743 |
+
"learning_rate": 4.320646572203033e-05,
|
| 1744 |
+
"loss": 1.0998,
|
| 1745 |
+
"step": 1235
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 0.7256866178669239,
|
| 1749 |
+
"grad_norm": 0.06982421875,
|
| 1750 |
+
"learning_rate": 4.236827495148443e-05,
|
| 1751 |
+
"loss": 1.0993,
|
| 1752 |
+
"step": 1240
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 0.7286127735841292,
|
| 1756 |
+
"grad_norm": 0.068359375,
|
| 1757 |
+
"learning_rate": 4.153610353456654e-05,
|
| 1758 |
+
"loss": 1.1323,
|
| 1759 |
+
"step": 1245
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 0.7315389293013346,
|
| 1763 |
+
"grad_norm": 0.06982421875,
|
| 1764 |
+
"learning_rate": 4.071003838753737e-05,
|
| 1765 |
+
"loss": 1.1264,
|
| 1766 |
+
"step": 1250
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 0.7344650850185399,
|
| 1770 |
+
"grad_norm": 0.0693359375,
|
| 1771 |
+
"learning_rate": 3.9890165788887365e-05,
|
| 1772 |
+
"loss": 1.1057,
|
| 1773 |
+
"step": 1255
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.7373912407357452,
|
| 1777 |
+
"grad_norm": 0.06884765625,
|
| 1778 |
+
"learning_rate": 3.9076571370325364e-05,
|
| 1779 |
+
"loss": 1.1119,
|
| 1780 |
+
"step": 1260
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.7403173964529506,
|
| 1784 |
+
"grad_norm": 0.0712890625,
|
| 1785 |
+
"learning_rate": 3.82693401078349e-05,
|
| 1786 |
+
"loss": 1.0996,
|
| 1787 |
+
"step": 1265
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.7432435521701559,
|
| 1791 |
+
"grad_norm": 0.06787109375,
|
| 1792 |
+
"learning_rate": 3.7468556312798685e-05,
|
| 1793 |
+
"loss": 1.1051,
|
| 1794 |
+
"step": 1270
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.7461697078873613,
|
| 1798 |
+
"grad_norm": 0.06787109375,
|
| 1799 |
+
"learning_rate": 3.667430362319277e-05,
|
| 1800 |
+
"loss": 1.0959,
|
| 1801 |
+
"step": 1275
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.7490958636045666,
|
| 1805 |
+
"grad_norm": 0.068359375,
|
| 1806 |
+
"learning_rate": 3.588666499485115e-05,
|
| 1807 |
+
"loss": 1.1129,
|
| 1808 |
+
"step": 1280
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.7520220193217719,
|
| 1812 |
+
"grad_norm": 0.0693359375,
|
| 1813 |
+
"learning_rate": 3.510572269280097e-05,
|
| 1814 |
+
"loss": 1.1184,
|
| 1815 |
+
"step": 1285
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.7549481750389773,
|
| 1819 |
+
"grad_norm": 0.06787109375,
|
| 1820 |
+
"learning_rate": 3.433155828267089e-05,
|
| 1821 |
+
"loss": 1.1003,
|
| 1822 |
+
"step": 1290
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.7578743307561826,
|
| 1826 |
+
"grad_norm": 0.06787109375,
|
| 1827 |
+
"learning_rate": 3.356425262217164e-05,
|
| 1828 |
+
"loss": 1.106,
|
| 1829 |
+
"step": 1295
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.7608004864733879,
|
| 1833 |
+
"grad_norm": 0.06787109375,
|
| 1834 |
+
"learning_rate": 3.280388585265075e-05,
|
| 1835 |
+
"loss": 1.1066,
|
| 1836 |
+
"step": 1300
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.7637266421905933,
|
| 1840 |
+
"grad_norm": 0.06689453125,
|
| 1841 |
+
"learning_rate": 3.205053739072248e-05,
|
| 1842 |
+
"loss": 1.1026,
|
| 1843 |
+
"step": 1305
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.7666527979077986,
|
| 1847 |
+
"grad_norm": 0.06787109375,
|
| 1848 |
+
"learning_rate": 3.130428591997282e-05,
|
| 1849 |
+
"loss": 1.1256,
|
| 1850 |
+
"step": 1310
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.769578953625004,
|
| 1854 |
+
"grad_norm": 0.0673828125,
|
| 1855 |
+
"learning_rate": 3.0565209382741664e-05,
|
| 1856 |
+
"loss": 1.1018,
|
| 1857 |
+
"step": 1315
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.7725051093422093,
|
| 1861 |
+
"grad_norm": 0.068359375,
|
| 1862 |
+
"learning_rate": 2.9833384971981838e-05,
|
| 1863 |
+
"loss": 1.1099,
|
| 1864 |
+
"step": 1320
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.7754312650594147,
|
| 1868 |
+
"grad_norm": 0.0673828125,
|
| 1869 |
+
"learning_rate": 2.9108889123196824e-05,
|
| 1870 |
+
"loss": 1.0995,
|
| 1871 |
+
"step": 1325
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.7783574207766201,
|
| 1875 |
+
"grad_norm": 0.0673828125,
|
| 1876 |
+
"learning_rate": 2.839179750645752e-05,
|
| 1877 |
+
"loss": 1.1194,
|
| 1878 |
+
"step": 1330
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.7812835764938254,
|
| 1882 |
+
"grad_norm": 0.0693359375,
|
| 1883 |
+
"learning_rate": 2.768218501849862e-05,
|
| 1884 |
+
"loss": 1.0955,
|
| 1885 |
+
"step": 1335
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.7842097322110307,
|
| 1889 |
+
"grad_norm": 0.06591796875,
|
| 1890 |
+
"learning_rate": 2.6980125774896238e-05,
|
| 1891 |
+
"loss": 1.0712,
|
| 1892 |
+
"step": 1340
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.7871358879282361,
|
| 1896 |
+
"grad_norm": 0.0693359375,
|
| 1897 |
+
"learning_rate": 2.6285693102326868e-05,
|
| 1898 |
+
"loss": 1.1019,
|
| 1899 |
+
"step": 1345
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.7900620436454414,
|
| 1903 |
+
"grad_norm": 0.0673828125,
|
| 1904 |
+
"learning_rate": 2.559895953090856e-05,
|
| 1905 |
+
"loss": 1.1022,
|
| 1906 |
+
"step": 1350
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.7929881993626468,
|
| 1910 |
+
"grad_norm": 0.06787109375,
|
| 1911 |
+
"learning_rate": 2.491999678662582e-05,
|
| 1912 |
+
"loss": 1.1027,
|
| 1913 |
+
"step": 1355
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.7959143550798521,
|
| 1917 |
+
"grad_norm": 0.06787109375,
|
| 1918 |
+
"learning_rate": 2.4248875783837987e-05,
|
| 1919 |
+
"loss": 1.1292,
|
| 1920 |
+
"step": 1360
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.7988405107970574,
|
| 1924 |
+
"grad_norm": 0.0693359375,
|
| 1925 |
+
"learning_rate": 2.358566661787257e-05,
|
| 1926 |
+
"loss": 1.1117,
|
| 1927 |
+
"step": 1365
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.8017666665142628,
|
| 1931 |
+
"grad_norm": 0.06787109375,
|
| 1932 |
+
"learning_rate": 2.293043855770416e-05,
|
| 1933 |
+
"loss": 1.1176,
|
| 1934 |
+
"step": 1370
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.8046928222314681,
|
| 1938 |
+
"grad_norm": 0.06982421875,
|
| 1939 |
+
"learning_rate": 2.2283260038719646e-05,
|
| 1940 |
+
"loss": 1.1074,
|
| 1941 |
+
"step": 1375
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.8076189779486734,
|
| 1945 |
+
"grad_norm": 0.06689453125,
|
| 1946 |
+
"learning_rate": 2.1644198655570504e-05,
|
| 1947 |
+
"loss": 1.1123,
|
| 1948 |
+
"step": 1380
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.8105451336658788,
|
| 1952 |
+
"grad_norm": 0.068359375,
|
| 1953 |
+
"learning_rate": 2.1013321155112754e-05,
|
| 1954 |
+
"loss": 1.0979,
|
| 1955 |
+
"step": 1385
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.8134712893830841,
|
| 1959 |
+
"grad_norm": 0.068359375,
|
| 1960 |
+
"learning_rate": 2.0390693429435627e-05,
|
| 1961 |
+
"loss": 1.1102,
|
| 1962 |
+
"step": 1390
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.8163974451002894,
|
| 1966 |
+
"grad_norm": 0.06884765625,
|
| 1967 |
+
"learning_rate": 1.977638050897954e-05,
|
| 1968 |
+
"loss": 1.1133,
|
| 1969 |
+
"step": 1395
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.8193236008174948,
|
| 1973 |
+
"grad_norm": 0.0673828125,
|
| 1974 |
+
"learning_rate": 1.917044655574387e-05,
|
| 1975 |
+
"loss": 1.1045,
|
| 1976 |
+
"step": 1400
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.8222497565347001,
|
| 1980 |
+
"grad_norm": 0.0673828125,
|
| 1981 |
+
"learning_rate": 1.8572954856585535e-05,
|
| 1982 |
+
"loss": 1.0967,
|
| 1983 |
+
"step": 1405
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.8251759122519055,
|
| 1987 |
+
"grad_norm": 0.0673828125,
|
| 1988 |
+
"learning_rate": 1.798396781660914e-05,
|
| 1989 |
+
"loss": 1.1199,
|
| 1990 |
+
"step": 1410
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.8281020679691108,
|
| 1994 |
+
"grad_norm": 0.06884765625,
|
| 1995 |
+
"learning_rate": 1.7403546952648885e-05,
|
| 1996 |
+
"loss": 1.1039,
|
| 1997 |
+
"step": 1415
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.8310282236863161,
|
| 2001 |
+
"grad_norm": 0.06787109375,
|
| 2002 |
+
"learning_rate": 1.6831752886843512e-05,
|
| 2003 |
+
"loss": 1.1106,
|
| 2004 |
+
"step": 1420
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.8339543794035215,
|
| 2008 |
+
"grad_norm": 0.06982421875,
|
| 2009 |
+
"learning_rate": 1.626864534030469e-05,
|
| 2010 |
+
"loss": 1.106,
|
| 2011 |
+
"step": 1425
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.8368805351207268,
|
| 2015 |
+
"grad_norm": 0.0673828125,
|
| 2016 |
+
"learning_rate": 1.571428312687928e-05,
|
| 2017 |
+
"loss": 1.1004,
|
| 2018 |
+
"step": 1430
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.8398066908379321,
|
| 2022 |
+
"grad_norm": 0.0693359375,
|
| 2023 |
+
"learning_rate": 1.5168724147006652e-05,
|
| 2024 |
+
"loss": 1.1244,
|
| 2025 |
+
"step": 1435
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.8427328465551375,
|
| 2029 |
+
"grad_norm": 0.0673828125,
|
| 2030 |
+
"learning_rate": 1.4632025381671133e-05,
|
| 2031 |
+
"loss": 1.1227,
|
| 2032 |
+
"step": 1440
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.8456590022723428,
|
| 2036 |
+
"grad_norm": 0.0673828125,
|
| 2037 |
+
"learning_rate": 1.4104242886450824e-05,
|
| 2038 |
+
"loss": 1.1073,
|
| 2039 |
+
"step": 1445
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.8485851579895481,
|
| 2043 |
+
"grad_norm": 0.07568359375,
|
| 2044 |
+
"learning_rate": 1.3585431785662627e-05,
|
| 2045 |
+
"loss": 1.0903,
|
| 2046 |
+
"step": 1450
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.8515113137067535,
|
| 2050 |
+
"grad_norm": 0.06884765625,
|
| 2051 |
+
"learning_rate": 1.3075646266604913e-05,
|
| 2052 |
+
"loss": 1.1129,
|
| 2053 |
+
"step": 1455
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.8544374694239588,
|
| 2057 |
+
"grad_norm": 0.06982421875,
|
| 2058 |
+
"learning_rate": 1.257493957389796e-05,
|
| 2059 |
+
"loss": 1.1293,
|
| 2060 |
+
"step": 1460
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 0.8573636251411642,
|
| 2064 |
+
"grad_norm": 0.0693359375,
|
| 2065 |
+
"learning_rate": 1.208336400392268e-05,
|
| 2066 |
+
"loss": 1.0987,
|
| 2067 |
+
"step": 1465
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.8602897808583695,
|
| 2071 |
+
"grad_norm": 0.06689453125,
|
| 2072 |
+
"learning_rate": 1.1600970899358588e-05,
|
| 2073 |
+
"loss": 1.1044,
|
| 2074 |
+
"step": 1470
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 0.8632159365755748,
|
| 2078 |
+
"grad_norm": 0.0693359375,
|
| 2079 |
+
"learning_rate": 1.1127810643821401e-05,
|
| 2080 |
+
"loss": 1.1182,
|
| 2081 |
+
"step": 1475
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 0.8661420922927802,
|
| 2085 |
+
"grad_norm": 0.0673828125,
|
| 2086 |
+
"learning_rate": 1.0663932656600505e-05,
|
| 2087 |
+
"loss": 1.0957,
|
| 2088 |
+
"step": 1480
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 0.8690682480099855,
|
| 2092 |
+
"grad_norm": 0.06689453125,
|
| 2093 |
+
"learning_rate": 1.0209385387497517e-05,
|
| 2094 |
+
"loss": 1.1238,
|
| 2095 |
+
"step": 1485
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 0.8719944037271908,
|
| 2099 |
+
"grad_norm": 0.068359375,
|
| 2100 |
+
"learning_rate": 9.764216311765905e-06,
|
| 2101 |
+
"loss": 1.1209,
|
| 2102 |
+
"step": 1490
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 0.8749205594443962,
|
| 2106 |
+
"grad_norm": 0.06591796875,
|
| 2107 |
+
"learning_rate": 9.328471925152381e-06,
|
| 2108 |
+
"loss": 1.1046,
|
| 2109 |
+
"step": 1495
|
| 2110 |
+
},
|
| 2111 |
+
{
|
| 2112 |
+
"epoch": 0.8778467151616015,
|
| 2113 |
+
"grad_norm": 0.06787109375,
|
| 2114 |
+
"learning_rate": 8.902197739040708e-06,
|
| 2115 |
+
"loss": 1.1205,
|
| 2116 |
+
"step": 1500
|
| 2117 |
+
},
|
| 2118 |
+
{
|
| 2119 |
+
"epoch": 0.8807728708788068,
|
| 2120 |
+
"grad_norm": 0.06787109375,
|
| 2121 |
+
"learning_rate": 8.485438275698154e-06,
|
| 2122 |
+
"loss": 1.1202,
|
| 2123 |
+
"step": 1505
|
| 2124 |
+
},
|
| 2125 |
+
{
|
| 2126 |
+
"epoch": 0.8836990265960122,
|
| 2127 |
+
"grad_norm": 0.07080078125,
|
| 2128 |
+
"learning_rate": 8.078237063625538e-06,
|
| 2129 |
+
"loss": 1.1177,
|
| 2130 |
+
"step": 1510
|
| 2131 |
+
},
|
| 2132 |
+
{
|
| 2133 |
+
"epoch": 0.8866251823132175,
|
| 2134 |
+
"grad_norm": 0.0673828125,
|
| 2135 |
+
"learning_rate": 7.680636633010695e-06,
|
| 2136 |
+
"loss": 1.1116,
|
| 2137 |
+
"step": 1515
|
| 2138 |
+
},
|
| 2139 |
+
{
|
| 2140 |
+
"epoch": 0.8895513380304229,
|
| 2141 |
+
"grad_norm": 0.0712890625,
|
| 2142 |
+
"learning_rate": 7.292678511286522e-06,
|
| 2143 |
+
"loss": 1.1067,
|
| 2144 |
+
"step": 1520
|
| 2145 |
+
},
|
| 2146 |
+
{
|
| 2147 |
+
"epoch": 0.8924774937476282,
|
| 2148 |
+
"grad_norm": 0.06884765625,
|
| 2149 |
+
"learning_rate": 6.914403218793608e-06,
|
| 2150 |
+
"loss": 1.12,
|
| 2151 |
+
"step": 1525
|
| 2152 |
+
},
|
| 2153 |
+
{
|
| 2154 |
+
"epoch": 0.8954036494648335,
|
| 2155 |
+
"grad_norm": 0.0693359375,
|
| 2156 |
+
"learning_rate": 6.5458502645480924e-06,
|
| 2157 |
+
"loss": 1.1298,
|
| 2158 |
+
"step": 1530
|
| 2159 |
+
},
|
| 2160 |
+
{
|
| 2161 |
+
"epoch": 0.8983298051820389,
|
| 2162 |
+
"grad_norm": 0.0673828125,
|
| 2163 |
+
"learning_rate": 6.187058142115077e-06,
|
| 2164 |
+
"loss": 1.1069,
|
| 2165 |
+
"step": 1535
|
| 2166 |
+
},
|
| 2167 |
+
{
|
| 2168 |
+
"epoch": 0.9012559608992442,
|
| 2169 |
+
"grad_norm": 0.06640625,
|
| 2170 |
+
"learning_rate": 5.838064325588288e-06,
|
| 2171 |
+
"loss": 1.0941,
|
| 2172 |
+
"step": 1540
|
| 2173 |
+
},
|
| 2174 |
+
{
|
| 2175 |
+
"epoch": 0.9041821166164495,
|
| 2176 |
+
"grad_norm": 0.06640625,
|
| 2177 |
+
"learning_rate": 5.498905265675958e-06,
|
| 2178 |
+
"loss": 1.0976,
|
| 2179 |
+
"step": 1545
|
| 2180 |
+
},
|
| 2181 |
+
{
|
| 2182 |
+
"epoch": 0.9071082723336549,
|
| 2183 |
+
"grad_norm": 0.06787109375,
|
| 2184 |
+
"learning_rate": 5.169616385893794e-06,
|
| 2185 |
+
"loss": 1.101,
|
| 2186 |
+
"step": 1550
|
| 2187 |
+
},
|
| 2188 |
+
{
|
| 2189 |
+
"epoch": 0.9100344280508602,
|
| 2190 |
+
"grad_norm": 0.0654296875,
|
| 2191 |
+
"learning_rate": 4.850232078865169e-06,
|
| 2192 |
+
"loss": 1.121,
|
| 2193 |
+
"step": 1555
|
| 2194 |
+
},
|
| 2195 |
+
{
|
| 2196 |
+
"epoch": 0.9129605837680655,
|
| 2197 |
+
"grad_norm": 0.068359375,
|
| 2198 |
+
"learning_rate": 4.5407857027289555e-06,
|
| 2199 |
+
"loss": 1.1013,
|
| 2200 |
+
"step": 1560
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 0.9158867394852709,
|
| 2204 |
+
"grad_norm": 0.0693359375,
|
| 2205 |
+
"learning_rate": 4.241309577655406e-06,
|
| 2206 |
+
"loss": 1.1464,
|
| 2207 |
+
"step": 1565
|
| 2208 |
+
},
|
| 2209 |
+
{
|
| 2210 |
+
"epoch": 0.9188128952024762,
|
| 2211 |
+
"grad_norm": 0.0654296875,
|
| 2212 |
+
"learning_rate": 3.951834982470526e-06,
|
| 2213 |
+
"loss": 1.1111,
|
| 2214 |
+
"step": 1570
|
| 2215 |
+
},
|
| 2216 |
+
{
|
| 2217 |
+
"epoch": 0.9217390509196816,
|
| 2218 |
+
"grad_norm": 0.06640625,
|
| 2219 |
+
"learning_rate": 3.672392151389137e-06,
|
| 2220 |
+
"loss": 1.1078,
|
| 2221 |
+
"step": 1575
|
| 2222 |
+
},
|
| 2223 |
+
{
|
| 2224 |
+
"epoch": 0.9246652066368869,
|
| 2225 |
+
"grad_norm": 0.07080078125,
|
| 2226 |
+
"learning_rate": 3.4030102708570212e-06,
|
| 2227 |
+
"loss": 1.1195,
|
| 2228 |
+
"step": 1580
|
| 2229 |
+
},
|
| 2230 |
+
{
|
| 2231 |
+
"epoch": 0.9275913623540922,
|
| 2232 |
+
"grad_norm": 0.0703125,
|
| 2233 |
+
"learning_rate": 3.143717476502572e-06,
|
| 2234 |
+
"loss": 1.1216,
|
| 2235 |
+
"step": 1585
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 0.9305175180712976,
|
| 2239 |
+
"grad_norm": 0.0673828125,
|
| 2240 |
+
"learning_rate": 2.8945408501981906e-06,
|
| 2241 |
+
"loss": 1.0936,
|
| 2242 |
+
"step": 1590
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 0.9334436737885029,
|
| 2246 |
+
"grad_norm": 0.06787109375,
|
| 2247 |
+
"learning_rate": 2.6555064172316234e-06,
|
| 2248 |
+
"loss": 1.106,
|
| 2249 |
+
"step": 1595
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 0.9363698295057082,
|
| 2253 |
+
"grad_norm": 0.068359375,
|
| 2254 |
+
"learning_rate": 2.4266391435878387e-06,
|
| 2255 |
+
"loss": 1.1061,
|
| 2256 |
+
"step": 1600
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 0.9392959852229136,
|
| 2260 |
+
"grad_norm": 0.068359375,
|
| 2261 |
+
"learning_rate": 2.2079629333414453e-06,
|
| 2262 |
+
"loss": 1.1119,
|
| 2263 |
+
"step": 1605
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"epoch": 0.9422221409401189,
|
| 2267 |
+
"grad_norm": 0.06787109375,
|
| 2268 |
+
"learning_rate": 1.999500626159967e-06,
|
| 2269 |
+
"loss": 1.1094,
|
| 2270 |
+
"step": 1610
|
| 2271 |
+
},
|
| 2272 |
+
{
|
| 2273 |
+
"epoch": 0.9451482966573242,
|
| 2274 |
+
"grad_norm": 0.06884765625,
|
| 2275 |
+
"learning_rate": 1.8012739949183844e-06,
|
| 2276 |
+
"loss": 1.1141,
|
| 2277 |
+
"step": 1615
|
| 2278 |
+
},
|
| 2279 |
+
{
|
| 2280 |
+
"epoch": 0.9480744523745297,
|
| 2281 |
+
"grad_norm": 0.06591796875,
|
| 2282 |
+
"learning_rate": 1.6133037434250985e-06,
|
| 2283 |
+
"loss": 1.1084,
|
| 2284 |
+
"step": 1620
|
| 2285 |
+
},
|
| 2286 |
+
{
|
| 2287 |
+
"epoch": 0.951000608091735,
|
| 2288 |
+
"grad_norm": 0.068359375,
|
| 2289 |
+
"learning_rate": 1.4356095042594386e-06,
|
| 2290 |
+
"loss": 1.1208,
|
| 2291 |
+
"step": 1625
|
| 2292 |
+
},
|
| 2293 |
+
{
|
| 2294 |
+
"epoch": 0.9539267638089404,
|
| 2295 |
+
"grad_norm": 0.0693359375,
|
| 2296 |
+
"learning_rate": 1.2682098367212237e-06,
|
| 2297 |
+
"loss": 1.1116,
|
| 2298 |
+
"step": 1630
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.9568529195261457,
|
| 2302 |
+
"grad_norm": 0.06640625,
|
| 2303 |
+
"learning_rate": 1.1111222248922471e-06,
|
| 2304 |
+
"loss": 1.1047,
|
| 2305 |
+
"step": 1635
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 0.959779075243351,
|
| 2309 |
+
"grad_norm": 0.0673828125,
|
| 2310 |
+
"learning_rate": 9.643630758102484e-07,
|
| 2311 |
+
"loss": 1.0998,
|
| 2312 |
+
"step": 1640
|
| 2313 |
+
},
|
| 2314 |
+
{
|
| 2315 |
+
"epoch": 0.9627052309605564,
|
| 2316 |
+
"grad_norm": 0.0654296875,
|
| 2317 |
+
"learning_rate": 8.279477177551842e-07,
|
| 2318 |
+
"loss": 1.1073,
|
| 2319 |
+
"step": 1645
|
| 2320 |
+
},
|
| 2321 |
+
{
|
| 2322 |
+
"epoch": 0.9656313866777617,
|
| 2323 |
+
"grad_norm": 0.0673828125,
|
| 2324 |
+
"learning_rate": 7.018903986483083e-07,
|
| 2325 |
+
"loss": 1.1124,
|
| 2326 |
+
"step": 1650
|
| 2327 |
+
},
|
| 2328 |
+
{
|
| 2329 |
+
"epoch": 0.968557542394967,
|
| 2330 |
+
"grad_norm": 0.0654296875,
|
| 2331 |
+
"learning_rate": 5.862042845640403e-07,
|
| 2332 |
+
"loss": 1.1023,
|
| 2333 |
+
"step": 1655
|
| 2334 |
+
},
|
| 2335 |
+
{
|
| 2336 |
+
"epoch": 0.9714836981121724,
|
| 2337 |
+
"grad_norm": 0.06396484375,
|
| 2338 |
+
"learning_rate": 4.809014583548432e-07,
|
| 2339 |
+
"loss": 1.1234,
|
| 2340 |
+
"step": 1660
|
| 2341 |
+
},
|
| 2342 |
+
{
|
| 2343 |
+
"epoch": 0.9744098538293777,
|
| 2344 |
+
"grad_norm": 0.06982421875,
|
| 2345 |
+
"learning_rate": 3.859929183892108e-07,
|
| 2346 |
+
"loss": 1.111,
|
| 2347 |
+
"step": 1665
|
| 2348 |
+
},
|
| 2349 |
+
{
|
| 2350 |
+
"epoch": 0.9773360095465831,
|
| 2351 |
+
"grad_norm": 0.06787109375,
|
| 2352 |
+
"learning_rate": 3.014885774029419e-07,
|
| 2353 |
+
"loss": 1.1179,
|
| 2354 |
+
"step": 1670
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.9802621652637884,
|
| 2358 |
+
"grad_norm": 0.064453125,
|
| 2359 |
+
"learning_rate": 2.2739726146381311e-07,
|
| 2360 |
+
"loss": 1.1082,
|
| 2361 |
+
"step": 1675
|
| 2362 |
+
},
|
| 2363 |
+
{
|
| 2364 |
+
"epoch": 0.9831883209809937,
|
| 2365 |
+
"grad_norm": 0.06982421875,
|
| 2366 |
+
"learning_rate": 1.6372670904974963e-07,
|
| 2367 |
+
"loss": 1.14,
|
| 2368 |
+
"step": 1680
|
| 2369 |
+
},
|
| 2370 |
+
{
|
| 2371 |
+
"epoch": 0.9861144766981991,
|
| 2372 |
+
"grad_norm": 0.06591796875,
|
| 2373 |
+
"learning_rate": 1.1048357024054934e-07,
|
| 2374 |
+
"loss": 1.1126,
|
| 2375 |
+
"step": 1685
|
| 2376 |
+
},
|
| 2377 |
+
{
|
| 2378 |
+
"epoch": 0.9890406324154044,
|
| 2379 |
+
"grad_norm": 0.06689453125,
|
| 2380 |
+
"learning_rate": 6.76734060233275e-08,
|
| 2381 |
+
"loss": 1.1087,
|
| 2382 |
+
"step": 1690
|
| 2383 |
+
},
|
| 2384 |
+
{
|
| 2385 |
+
"epoch": 0.9919667881326097,
|
| 2386 |
+
"grad_norm": 0.06591796875,
|
| 2387 |
+
"learning_rate": 3.5300687711703475e-08,
|
| 2388 |
+
"loss": 1.1235,
|
| 2389 |
+
"step": 1695
|
| 2390 |
+
},
|
| 2391 |
+
{
|
| 2392 |
+
"epoch": 0.9948929438498151,
|
| 2393 |
+
"grad_norm": 0.0654296875,
|
| 2394 |
+
"learning_rate": 1.3368796478807621e-08,
|
| 2395 |
+
"loss": 1.1019,
|
| 2396 |
+
"step": 1700
|
| 2397 |
+
},
|
| 2398 |
+
{
|
| 2399 |
+
"epoch": 0.9978190995670204,
|
| 2400 |
+
"grad_norm": 0.06787109375,
|
| 2401 |
+
"learning_rate": 1.8800230040860733e-09,
|
| 2402 |
+
"loss": 1.1143,
|
| 2403 |
+
"step": 1705
|
| 2404 |
+
},
|
| 2405 |
+
{
|
| 2406 |
+
"epoch": 0.9995747929973436,
|
| 2407 |
+
"eval_loss": 1.278271198272705,
|
| 2408 |
+
"eval_runtime": 1249.6014,
|
| 2409 |
+
"eval_samples_per_second": 12.937,
|
| 2410 |
+
"eval_steps_per_second": 12.937,
|
| 2411 |
+
"step": 1708
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.9995747929973436,
|
| 2415 |
+
"step": 1708,
|
| 2416 |
+
"total_flos": 2.8130589802160456e+18,
|
| 2417 |
+
"train_loss": 0.978924176871637,
|
| 2418 |
+
"train_runtime": 53778.1468,
|
| 2419 |
+
"train_samples_per_second": 4.067,
|
| 2420 |
+
"train_steps_per_second": 0.032
|
| 2421 |
+
}
|
| 2422 |
+
],
|
| 2423 |
+
"logging_steps": 5,
|
| 2424 |
+
"max_steps": 1708,
|
| 2425 |
+
"num_input_tokens_seen": 0,
|
| 2426 |
+
"num_train_epochs": 1,
|
| 2427 |
+
"save_steps": 20,
|
| 2428 |
+
"total_flos": 2.8130589802160456e+18,
|
| 2429 |
+
"train_batch_size": 1,
|
| 2430 |
+
"trial_name": null,
|
| 2431 |
+
"trial_params": null
|
| 2432 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:82ba869b649e48186d8305900224ddb4c5e43edf1597dc49af2bc54ef7f618ae
|
| 3 |
+
size 5112
|