Training in progress, step 375, checkpoint
Browse files- checkpoint-375/README.md +202 -0
- checkpoint-375/adapter_config.json +37 -0
- checkpoint-375/adapter_model.safetensors +3 -0
- checkpoint-375/global_step375/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- checkpoint-375/global_step375/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- checkpoint-375/global_step375/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- checkpoint-375/global_step375/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
- checkpoint-375/global_step375/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- checkpoint-375/global_step375/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
- checkpoint-375/global_step375/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
- checkpoint-375/global_step375/zero_pp_rank_3_mp_rank_00_model_states.pt +3 -0
- checkpoint-375/latest +1 -0
- checkpoint-375/rng_state_0.pth +3 -0
- checkpoint-375/rng_state_1.pth +3 -0
- checkpoint-375/rng_state_2.pth +3 -0
- checkpoint-375/rng_state_3.pth +3 -0
- checkpoint-375/scheduler.pt +3 -0
- checkpoint-375/special_tokens_map.json +54 -0
- checkpoint-375/tokenizer.model +3 -0
- checkpoint-375/tokenizer_config.json +249 -0
- checkpoint-375/trainer_state.json +2706 -0
- checkpoint-375/training_args.bin +3 -0
- checkpoint-375/zero_to_fp32.py +760 -0
checkpoint-375/README.md
ADDED
|
@@ -0,0 +1,202 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: internlm/internlm3-8b-instruct
|
| 3 |
+
library_name: peft
|
| 4 |
+
---
|
| 5 |
+
|
| 6 |
+
# Model Card for Model ID
|
| 7 |
+
|
| 8 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
|
| 12 |
+
## Model Details
|
| 13 |
+
|
| 14 |
+
### Model Description
|
| 15 |
+
|
| 16 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
|
| 20 |
+
- **Developed by:** [More Information Needed]
|
| 21 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 22 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 23 |
+
- **Model type:** [More Information Needed]
|
| 24 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 25 |
+
- **License:** [More Information Needed]
|
| 26 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 27 |
+
|
| 28 |
+
### Model Sources [optional]
|
| 29 |
+
|
| 30 |
+
<!-- Provide the basic links for the model. -->
|
| 31 |
+
|
| 32 |
+
- **Repository:** [More Information Needed]
|
| 33 |
+
- **Paper [optional]:** [More Information Needed]
|
| 34 |
+
- **Demo [optional]:** [More Information Needed]
|
| 35 |
+
|
| 36 |
+
## Uses
|
| 37 |
+
|
| 38 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 39 |
+
|
| 40 |
+
### Direct Use
|
| 41 |
+
|
| 42 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 43 |
+
|
| 44 |
+
[More Information Needed]
|
| 45 |
+
|
| 46 |
+
### Downstream Use [optional]
|
| 47 |
+
|
| 48 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 49 |
+
|
| 50 |
+
[More Information Needed]
|
| 51 |
+
|
| 52 |
+
### Out-of-Scope Use
|
| 53 |
+
|
| 54 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 55 |
+
|
| 56 |
+
[More Information Needed]
|
| 57 |
+
|
| 58 |
+
## Bias, Risks, and Limitations
|
| 59 |
+
|
| 60 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 61 |
+
|
| 62 |
+
[More Information Needed]
|
| 63 |
+
|
| 64 |
+
### Recommendations
|
| 65 |
+
|
| 66 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 67 |
+
|
| 68 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 69 |
+
|
| 70 |
+
## How to Get Started with the Model
|
| 71 |
+
|
| 72 |
+
Use the code below to get started with the model.
|
| 73 |
+
|
| 74 |
+
[More Information Needed]
|
| 75 |
+
|
| 76 |
+
## Training Details
|
| 77 |
+
|
| 78 |
+
### Training Data
|
| 79 |
+
|
| 80 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 81 |
+
|
| 82 |
+
[More Information Needed]
|
| 83 |
+
|
| 84 |
+
### Training Procedure
|
| 85 |
+
|
| 86 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 87 |
+
|
| 88 |
+
#### Preprocessing [optional]
|
| 89 |
+
|
| 90 |
+
[More Information Needed]
|
| 91 |
+
|
| 92 |
+
|
| 93 |
+
#### Training Hyperparameters
|
| 94 |
+
|
| 95 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 96 |
+
|
| 97 |
+
#### Speeds, Sizes, Times [optional]
|
| 98 |
+
|
| 99 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 100 |
+
|
| 101 |
+
[More Information Needed]
|
| 102 |
+
|
| 103 |
+
## Evaluation
|
| 104 |
+
|
| 105 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 106 |
+
|
| 107 |
+
### Testing Data, Factors & Metrics
|
| 108 |
+
|
| 109 |
+
#### Testing Data
|
| 110 |
+
|
| 111 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 112 |
+
|
| 113 |
+
[More Information Needed]
|
| 114 |
+
|
| 115 |
+
#### Factors
|
| 116 |
+
|
| 117 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 118 |
+
|
| 119 |
+
[More Information Needed]
|
| 120 |
+
|
| 121 |
+
#### Metrics
|
| 122 |
+
|
| 123 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 124 |
+
|
| 125 |
+
[More Information Needed]
|
| 126 |
+
|
| 127 |
+
### Results
|
| 128 |
+
|
| 129 |
+
[More Information Needed]
|
| 130 |
+
|
| 131 |
+
#### Summary
|
| 132 |
+
|
| 133 |
+
|
| 134 |
+
|
| 135 |
+
## Model Examination [optional]
|
| 136 |
+
|
| 137 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 138 |
+
|
| 139 |
+
[More Information Needed]
|
| 140 |
+
|
| 141 |
+
## Environmental Impact
|
| 142 |
+
|
| 143 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 144 |
+
|
| 145 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 146 |
+
|
| 147 |
+
- **Hardware Type:** [More Information Needed]
|
| 148 |
+
- **Hours used:** [More Information Needed]
|
| 149 |
+
- **Cloud Provider:** [More Information Needed]
|
| 150 |
+
- **Compute Region:** [More Information Needed]
|
| 151 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 152 |
+
|
| 153 |
+
## Technical Specifications [optional]
|
| 154 |
+
|
| 155 |
+
### Model Architecture and Objective
|
| 156 |
+
|
| 157 |
+
[More Information Needed]
|
| 158 |
+
|
| 159 |
+
### Compute Infrastructure
|
| 160 |
+
|
| 161 |
+
[More Information Needed]
|
| 162 |
+
|
| 163 |
+
#### Hardware
|
| 164 |
+
|
| 165 |
+
[More Information Needed]
|
| 166 |
+
|
| 167 |
+
#### Software
|
| 168 |
+
|
| 169 |
+
[More Information Needed]
|
| 170 |
+
|
| 171 |
+
## Citation [optional]
|
| 172 |
+
|
| 173 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 174 |
+
|
| 175 |
+
**BibTeX:**
|
| 176 |
+
|
| 177 |
+
[More Information Needed]
|
| 178 |
+
|
| 179 |
+
**APA:**
|
| 180 |
+
|
| 181 |
+
[More Information Needed]
|
| 182 |
+
|
| 183 |
+
## Glossary [optional]
|
| 184 |
+
|
| 185 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 186 |
+
|
| 187 |
+
[More Information Needed]
|
| 188 |
+
|
| 189 |
+
## More Information [optional]
|
| 190 |
+
|
| 191 |
+
[More Information Needed]
|
| 192 |
+
|
| 193 |
+
## Model Card Authors [optional]
|
| 194 |
+
|
| 195 |
+
[More Information Needed]
|
| 196 |
+
|
| 197 |
+
## Model Card Contact
|
| 198 |
+
|
| 199 |
+
[More Information Needed]
|
| 200 |
+
### Framework versions
|
| 201 |
+
|
| 202 |
+
- PEFT 0.14.0
|
checkpoint-375/adapter_config.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alpha_pattern": {},
|
| 3 |
+
"auto_mapping": null,
|
| 4 |
+
"base_model_name_or_path": "internlm/internlm3-8b-instruct",
|
| 5 |
+
"bias": "none",
|
| 6 |
+
"eva_config": null,
|
| 7 |
+
"exclude_modules": null,
|
| 8 |
+
"fan_in_fan_out": null,
|
| 9 |
+
"inference_mode": true,
|
| 10 |
+
"init_lora_weights": true,
|
| 11 |
+
"layer_replication": null,
|
| 12 |
+
"layers_pattern": null,
|
| 13 |
+
"layers_to_transform": null,
|
| 14 |
+
"loftq_config": {},
|
| 15 |
+
"lora_alpha": 64,
|
| 16 |
+
"lora_bias": false,
|
| 17 |
+
"lora_dropout": 0.25,
|
| 18 |
+
"megatron_config": null,
|
| 19 |
+
"megatron_core": "megatron.core",
|
| 20 |
+
"modules_to_save": null,
|
| 21 |
+
"peft_type": "LORA",
|
| 22 |
+
"r": 32,
|
| 23 |
+
"rank_pattern": {},
|
| 24 |
+
"revision": null,
|
| 25 |
+
"target_modules": [
|
| 26 |
+
"v_proj",
|
| 27 |
+
"gate_proj",
|
| 28 |
+
"down_proj",
|
| 29 |
+
"k_proj",
|
| 30 |
+
"o_proj",
|
| 31 |
+
"q_proj",
|
| 32 |
+
"up_proj"
|
| 33 |
+
],
|
| 34 |
+
"task_type": "CAUSAL_LM",
|
| 35 |
+
"use_dora": false,
|
| 36 |
+
"use_rslora": false
|
| 37 |
+
}
|
checkpoint-375/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d8d71366d9cc2c09a22a5d09e4263c8cfef955840bdc71971d5e6de4b4020cb5
|
| 3 |
+
size 2308615184
|
checkpoint-375/global_step375/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6777fe9d01b773c5c60c8dc3e3fdfe2c58ffb7a9dee929d89c9ae3b374df12ba
|
| 3 |
+
size 187091776
|
checkpoint-375/global_step375/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3650c45007155caeb23626f5d6df1ff136ed1c1b430af95edd58c9208cb3fa81
|
| 3 |
+
size 187091776
|
checkpoint-375/global_step375/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bcb2f8cccd870d4680b92794df34756848bae4a73068b88a5012385e2643be41
|
| 3 |
+
size 187091776
|
checkpoint-375/global_step375/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9bbc96108858eaf05b95a352784a9937a3080058f587c4ef55def7179061ca64
|
| 3 |
+
size 187091776
|
checkpoint-375/global_step375/zero_pp_rank_0_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:24bbd85b65d42346d14b87b9605b56bf4d0a64a6cfdee8b8a89e0dacd6d58a86
|
| 3 |
+
size 124777254
|
checkpoint-375/global_step375/zero_pp_rank_1_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:abf6cf8650efb42b9c9205db6d0db0cdad598ef42df5d881699f6afc58d11004
|
| 3 |
+
size 124777254
|
checkpoint-375/global_step375/zero_pp_rank_2_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8b18c1501f4876f1c514b39a2c637247743cf5d0104ca24c1ca840b7e3fa49da
|
| 3 |
+
size 124777254
|
checkpoint-375/global_step375/zero_pp_rank_3_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:af1666d267f29623ce4468b4a7a58e795135ae9d42ec798a3079fd4a6b939c97
|
| 3 |
+
size 124777254
|
checkpoint-375/latest
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
global_step375
|
checkpoint-375/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db27ed1d9351ecef9f6fa8e0a0344db373dcd44e5cf66868a18b1d3d051d42b8
|
| 3 |
+
size 14960
|
checkpoint-375/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:326e222cf3fe2d14248046ad69a2c46c167e8e958b2e42179c6d85edb333ee49
|
| 3 |
+
size 14960
|
checkpoint-375/rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ace3ace7aec43fd1c6df07e0ad33bf89240b4b73fb9ea5d400f0b3191b943177
|
| 3 |
+
size 14960
|
checkpoint-375/rng_state_3.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0564fb18fa63fc0cc5d8a09463435503edb50ae52cc139d2ee5292cbba60a6bc
|
| 3 |
+
size 14960
|
checkpoint-375/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e8365a1287812bae366360ab7bc14486613ac684c2699e0c42d74f156faff0bd
|
| 3 |
+
size 1064
|
checkpoint-375/special_tokens_map.json
ADDED
|
@@ -0,0 +1,54 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>",
|
| 5 |
+
"<|action_start|>",
|
| 6 |
+
"<|action_end|>",
|
| 7 |
+
"<|interpreter|>",
|
| 8 |
+
"<|plugin|>",
|
| 9 |
+
"<restate>",
|
| 10 |
+
"</restate>",
|
| 11 |
+
"<planning>",
|
| 12 |
+
"</planning>",
|
| 13 |
+
"<recollect>",
|
| 14 |
+
"</recollect>",
|
| 15 |
+
"<execution>",
|
| 16 |
+
"</execution>",
|
| 17 |
+
"<review>",
|
| 18 |
+
"</review>",
|
| 19 |
+
"<summarize>",
|
| 20 |
+
"</summarize>",
|
| 21 |
+
"<retry>",
|
| 22 |
+
"</retry>",
|
| 23 |
+
"<conclude>",
|
| 24 |
+
"</conclude>"
|
| 25 |
+
],
|
| 26 |
+
"bos_token": {
|
| 27 |
+
"content": "<s>",
|
| 28 |
+
"lstrip": false,
|
| 29 |
+
"normalized": false,
|
| 30 |
+
"rstrip": false,
|
| 31 |
+
"single_word": false
|
| 32 |
+
},
|
| 33 |
+
"eos_token": {
|
| 34 |
+
"content": "</s>",
|
| 35 |
+
"lstrip": false,
|
| 36 |
+
"normalized": false,
|
| 37 |
+
"rstrip": false,
|
| 38 |
+
"single_word": false
|
| 39 |
+
},
|
| 40 |
+
"pad_token": {
|
| 41 |
+
"content": "</s>",
|
| 42 |
+
"lstrip": false,
|
| 43 |
+
"normalized": false,
|
| 44 |
+
"rstrip": false,
|
| 45 |
+
"single_word": false
|
| 46 |
+
},
|
| 47 |
+
"unk_token": {
|
| 48 |
+
"content": "<unk>",
|
| 49 |
+
"lstrip": false,
|
| 50 |
+
"normalized": false,
|
| 51 |
+
"rstrip": false,
|
| 52 |
+
"single_word": false
|
| 53 |
+
}
|
| 54 |
+
}
|
checkpoint-375/tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bcacff3229854f5103ee7a85473a30ca9a8b3a68f3aae9b7479574b23ac2256b
|
| 3 |
+
size 2475075
|
checkpoint-375/tokenizer_config.json
ADDED
|
@@ -0,0 +1,249 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": true,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
},
|
| 30 |
+
"128111": {
|
| 31 |
+
"content": "<restate>",
|
| 32 |
+
"lstrip": false,
|
| 33 |
+
"normalized": false,
|
| 34 |
+
"rstrip": false,
|
| 35 |
+
"single_word": false,
|
| 36 |
+
"special": true
|
| 37 |
+
},
|
| 38 |
+
"128112": {
|
| 39 |
+
"content": "</restate>",
|
| 40 |
+
"lstrip": false,
|
| 41 |
+
"normalized": false,
|
| 42 |
+
"rstrip": false,
|
| 43 |
+
"single_word": false,
|
| 44 |
+
"special": true
|
| 45 |
+
},
|
| 46 |
+
"128113": {
|
| 47 |
+
"content": "<planning>",
|
| 48 |
+
"lstrip": false,
|
| 49 |
+
"normalized": false,
|
| 50 |
+
"rstrip": false,
|
| 51 |
+
"single_word": false,
|
| 52 |
+
"special": true
|
| 53 |
+
},
|
| 54 |
+
"128114": {
|
| 55 |
+
"content": "</planning>",
|
| 56 |
+
"lstrip": false,
|
| 57 |
+
"normalized": false,
|
| 58 |
+
"rstrip": false,
|
| 59 |
+
"single_word": false,
|
| 60 |
+
"special": true
|
| 61 |
+
},
|
| 62 |
+
"128115": {
|
| 63 |
+
"content": "<recollect>",
|
| 64 |
+
"lstrip": false,
|
| 65 |
+
"normalized": false,
|
| 66 |
+
"rstrip": false,
|
| 67 |
+
"single_word": false,
|
| 68 |
+
"special": true
|
| 69 |
+
},
|
| 70 |
+
"128116": {
|
| 71 |
+
"content": "</recollect>",
|
| 72 |
+
"lstrip": false,
|
| 73 |
+
"normalized": false,
|
| 74 |
+
"rstrip": false,
|
| 75 |
+
"single_word": false,
|
| 76 |
+
"special": true
|
| 77 |
+
},
|
| 78 |
+
"128117": {
|
| 79 |
+
"content": "<execution>",
|
| 80 |
+
"lstrip": false,
|
| 81 |
+
"normalized": false,
|
| 82 |
+
"rstrip": false,
|
| 83 |
+
"single_word": false,
|
| 84 |
+
"special": true
|
| 85 |
+
},
|
| 86 |
+
"128118": {
|
| 87 |
+
"content": "</execution>",
|
| 88 |
+
"lstrip": false,
|
| 89 |
+
"normalized": false,
|
| 90 |
+
"rstrip": false,
|
| 91 |
+
"single_word": false,
|
| 92 |
+
"special": true
|
| 93 |
+
},
|
| 94 |
+
"128119": {
|
| 95 |
+
"content": "<review>",
|
| 96 |
+
"lstrip": false,
|
| 97 |
+
"normalized": false,
|
| 98 |
+
"rstrip": false,
|
| 99 |
+
"single_word": false,
|
| 100 |
+
"special": true
|
| 101 |
+
},
|
| 102 |
+
"128120": {
|
| 103 |
+
"content": "</review>",
|
| 104 |
+
"lstrip": false,
|
| 105 |
+
"normalized": false,
|
| 106 |
+
"rstrip": false,
|
| 107 |
+
"single_word": false,
|
| 108 |
+
"special": true
|
| 109 |
+
},
|
| 110 |
+
"128121": {
|
| 111 |
+
"content": "<summarize>",
|
| 112 |
+
"lstrip": false,
|
| 113 |
+
"normalized": false,
|
| 114 |
+
"rstrip": false,
|
| 115 |
+
"single_word": false,
|
| 116 |
+
"special": true
|
| 117 |
+
},
|
| 118 |
+
"128122": {
|
| 119 |
+
"content": "</summarize>",
|
| 120 |
+
"lstrip": false,
|
| 121 |
+
"normalized": false,
|
| 122 |
+
"rstrip": false,
|
| 123 |
+
"single_word": false,
|
| 124 |
+
"special": true
|
| 125 |
+
},
|
| 126 |
+
"128123": {
|
| 127 |
+
"content": "<retry>",
|
| 128 |
+
"lstrip": false,
|
| 129 |
+
"normalized": false,
|
| 130 |
+
"rstrip": false,
|
| 131 |
+
"single_word": false,
|
| 132 |
+
"special": true
|
| 133 |
+
},
|
| 134 |
+
"128124": {
|
| 135 |
+
"content": "</retry>",
|
| 136 |
+
"lstrip": false,
|
| 137 |
+
"normalized": false,
|
| 138 |
+
"rstrip": false,
|
| 139 |
+
"single_word": false,
|
| 140 |
+
"special": true
|
| 141 |
+
},
|
| 142 |
+
"128125": {
|
| 143 |
+
"content": "<conclude>",
|
| 144 |
+
"lstrip": false,
|
| 145 |
+
"normalized": false,
|
| 146 |
+
"rstrip": false,
|
| 147 |
+
"single_word": false,
|
| 148 |
+
"special": true
|
| 149 |
+
},
|
| 150 |
+
"128126": {
|
| 151 |
+
"content": "</conclude>",
|
| 152 |
+
"lstrip": false,
|
| 153 |
+
"normalized": false,
|
| 154 |
+
"rstrip": false,
|
| 155 |
+
"single_word": false,
|
| 156 |
+
"special": true
|
| 157 |
+
},
|
| 158 |
+
"128127": {
|
| 159 |
+
"content": "<|plugin|>",
|
| 160 |
+
"lstrip": false,
|
| 161 |
+
"normalized": false,
|
| 162 |
+
"rstrip": false,
|
| 163 |
+
"single_word": false,
|
| 164 |
+
"special": true
|
| 165 |
+
},
|
| 166 |
+
"128128": {
|
| 167 |
+
"content": "<|interpreter|>",
|
| 168 |
+
"lstrip": false,
|
| 169 |
+
"normalized": false,
|
| 170 |
+
"rstrip": false,
|
| 171 |
+
"single_word": false,
|
| 172 |
+
"special": true
|
| 173 |
+
},
|
| 174 |
+
"128129": {
|
| 175 |
+
"content": "<|action_end|>",
|
| 176 |
+
"lstrip": false,
|
| 177 |
+
"normalized": false,
|
| 178 |
+
"rstrip": false,
|
| 179 |
+
"single_word": false,
|
| 180 |
+
"special": true
|
| 181 |
+
},
|
| 182 |
+
"128130": {
|
| 183 |
+
"content": "<|action_start|>",
|
| 184 |
+
"lstrip": false,
|
| 185 |
+
"normalized": false,
|
| 186 |
+
"rstrip": false,
|
| 187 |
+
"single_word": false,
|
| 188 |
+
"special": true
|
| 189 |
+
},
|
| 190 |
+
"128131": {
|
| 191 |
+
"content": "<|im_end|>",
|
| 192 |
+
"lstrip": false,
|
| 193 |
+
"normalized": false,
|
| 194 |
+
"rstrip": false,
|
| 195 |
+
"single_word": false,
|
| 196 |
+
"special": true
|
| 197 |
+
},
|
| 198 |
+
"128132": {
|
| 199 |
+
"content": "<|im_start|>",
|
| 200 |
+
"lstrip": false,
|
| 201 |
+
"normalized": false,
|
| 202 |
+
"rstrip": false,
|
| 203 |
+
"single_word": false,
|
| 204 |
+
"special": true
|
| 205 |
+
}
|
| 206 |
+
},
|
| 207 |
+
"additional_special_tokens": [
|
| 208 |
+
"<|im_start|>",
|
| 209 |
+
"<|im_end|>",
|
| 210 |
+
"<|action_start|>",
|
| 211 |
+
"<|action_end|>",
|
| 212 |
+
"<|interpreter|>",
|
| 213 |
+
"<|plugin|>",
|
| 214 |
+
"<restate>",
|
| 215 |
+
"</restate>",
|
| 216 |
+
"<planning>",
|
| 217 |
+
"</planning>",
|
| 218 |
+
"<recollect>",
|
| 219 |
+
"</recollect>",
|
| 220 |
+
"<execution>",
|
| 221 |
+
"</execution>",
|
| 222 |
+
"<review>",
|
| 223 |
+
"</review>",
|
| 224 |
+
"<summarize>",
|
| 225 |
+
"</summarize>",
|
| 226 |
+
"<retry>",
|
| 227 |
+
"</retry>",
|
| 228 |
+
"<conclude>",
|
| 229 |
+
"</conclude>"
|
| 230 |
+
],
|
| 231 |
+
"auto_map": {
|
| 232 |
+
"AutoTokenizer": [
|
| 233 |
+
"internlm/internlm3-8b-instruct--tokenization_internlm3.InternLM3Tokenizer",
|
| 234 |
+
null
|
| 235 |
+
]
|
| 236 |
+
},
|
| 237 |
+
"bos_token": "<s>",
|
| 238 |
+
"chat_template": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
|
| 239 |
+
"clean_up_tokenization_spaces": false,
|
| 240 |
+
"eos_token": "</s>",
|
| 241 |
+
"extra_special_tokens": {},
|
| 242 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 243 |
+
"pad_token": "</s>",
|
| 244 |
+
"sp_model_kwargs": {},
|
| 245 |
+
"spaces_between_special_tokens": false,
|
| 246 |
+
"tokenizer_class": "InternLM3Tokenizer",
|
| 247 |
+
"unk_token": "<unk>",
|
| 248 |
+
"use_default_system_prompt": false
|
| 249 |
+
}
|
checkpoint-375/trainer_state.json
ADDED
|
@@ -0,0 +1,2706 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5,
|
| 5 |
+
"eval_steps": 75,
|
| 6 |
+
"global_step": 375,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.0013333333333333333,
|
| 13 |
+
"grad_norm": 5.494853571957073,
|
| 14 |
+
"learning_rate": 1.5e-06,
|
| 15 |
+
"loss": 2.2794,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.0013333333333333333,
|
| 20 |
+
"eval_loss": 1.8317261934280396,
|
| 21 |
+
"eval_runtime": 98.252,
|
| 22 |
+
"eval_samples_per_second": 1.018,
|
| 23 |
+
"eval_steps_per_second": 0.254,
|
| 24 |
+
"step": 1
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0026666666666666666,
|
| 28 |
+
"grad_norm": 0.4396391591358924,
|
| 29 |
+
"learning_rate": 3e-06,
|
| 30 |
+
"loss": 2.2089,
|
| 31 |
+
"step": 2
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.004,
|
| 35 |
+
"grad_norm": 0.6704432458412378,
|
| 36 |
+
"learning_rate": 4.5e-06,
|
| 37 |
+
"loss": 2.3789,
|
| 38 |
+
"step": 3
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.005333333333333333,
|
| 42 |
+
"grad_norm": 0.5524719828039701,
|
| 43 |
+
"learning_rate": 6e-06,
|
| 44 |
+
"loss": 1.9508,
|
| 45 |
+
"step": 4
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.006666666666666667,
|
| 49 |
+
"grad_norm": 1.551345196476793,
|
| 50 |
+
"learning_rate": 7.5e-06,
|
| 51 |
+
"loss": 2.3302,
|
| 52 |
+
"step": 5
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.008,
|
| 56 |
+
"grad_norm": 1.4425623787692365,
|
| 57 |
+
"learning_rate": 9e-06,
|
| 58 |
+
"loss": 2.4649,
|
| 59 |
+
"step": 6
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.009333333333333334,
|
| 63 |
+
"grad_norm": 1.9558403413673842,
|
| 64 |
+
"learning_rate": 1.05e-05,
|
| 65 |
+
"loss": 2.5318,
|
| 66 |
+
"step": 7
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.010666666666666666,
|
| 70 |
+
"grad_norm": 0.6930931382639645,
|
| 71 |
+
"learning_rate": 1.2e-05,
|
| 72 |
+
"loss": 1.9398,
|
| 73 |
+
"step": 8
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.012,
|
| 77 |
+
"grad_norm": 0.40522998481846323,
|
| 78 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 79 |
+
"loss": 2.4355,
|
| 80 |
+
"step": 9
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.013333333333333334,
|
| 84 |
+
"grad_norm": 2.194079323679347,
|
| 85 |
+
"learning_rate": 1.5e-05,
|
| 86 |
+
"loss": 2.3417,
|
| 87 |
+
"step": 10
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.014666666666666666,
|
| 91 |
+
"grad_norm": 2.5436460717739156,
|
| 92 |
+
"learning_rate": 1.65e-05,
|
| 93 |
+
"loss": 2.4571,
|
| 94 |
+
"step": 11
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.016,
|
| 98 |
+
"grad_norm": 1.9496092240793341,
|
| 99 |
+
"learning_rate": 1.8e-05,
|
| 100 |
+
"loss": 1.9637,
|
| 101 |
+
"step": 12
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.017333333333333333,
|
| 105 |
+
"grad_norm": 1.2616722727237577,
|
| 106 |
+
"learning_rate": 1.95e-05,
|
| 107 |
+
"loss": 2.5866,
|
| 108 |
+
"step": 13
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.018666666666666668,
|
| 112 |
+
"grad_norm": 0.3060862022495338,
|
| 113 |
+
"learning_rate": 2.1e-05,
|
| 114 |
+
"loss": 2.1547,
|
| 115 |
+
"step": 14
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.02,
|
| 119 |
+
"grad_norm": 0.5028816367161089,
|
| 120 |
+
"learning_rate": 2.25e-05,
|
| 121 |
+
"loss": 1.8957,
|
| 122 |
+
"step": 15
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.021333333333333333,
|
| 126 |
+
"grad_norm": 0.5034069784636721,
|
| 127 |
+
"learning_rate": 2.4e-05,
|
| 128 |
+
"loss": 2.1151,
|
| 129 |
+
"step": 16
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.02266666666666667,
|
| 133 |
+
"grad_norm": 1.5564960276262187,
|
| 134 |
+
"learning_rate": 2.55e-05,
|
| 135 |
+
"loss": 2.3041,
|
| 136 |
+
"step": 17
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.024,
|
| 140 |
+
"grad_norm": 0.4958603020666427,
|
| 141 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 142 |
+
"loss": 2.0962,
|
| 143 |
+
"step": 18
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.025333333333333333,
|
| 147 |
+
"grad_norm": 1.5362995691059922,
|
| 148 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 149 |
+
"loss": 2.3989,
|
| 150 |
+
"step": 19
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.02666666666666667,
|
| 154 |
+
"grad_norm": 0.648652241685575,
|
| 155 |
+
"learning_rate": 3e-05,
|
| 156 |
+
"loss": 2.5339,
|
| 157 |
+
"step": 20
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.028,
|
| 161 |
+
"grad_norm": 0.554470724602306,
|
| 162 |
+
"learning_rate": 2.999996958555301e-05,
|
| 163 |
+
"loss": 2.4793,
|
| 164 |
+
"step": 21
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.029333333333333333,
|
| 168 |
+
"grad_norm": 1.3508987707917082,
|
| 169 |
+
"learning_rate": 2.999987834234907e-05,
|
| 170 |
+
"loss": 2.3167,
|
| 171 |
+
"step": 22
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.030666666666666665,
|
| 175 |
+
"grad_norm": 0.4599504691462346,
|
| 176 |
+
"learning_rate": 2.9999726270799325e-05,
|
| 177 |
+
"loss": 2.36,
|
| 178 |
+
"step": 23
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.032,
|
| 182 |
+
"grad_norm": 1.0007538870075432,
|
| 183 |
+
"learning_rate": 2.999951337158897e-05,
|
| 184 |
+
"loss": 2.2342,
|
| 185 |
+
"step": 24
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.03333333333333333,
|
| 189 |
+
"grad_norm": 0.8436818785690442,
|
| 190 |
+
"learning_rate": 2.9999239645677304e-05,
|
| 191 |
+
"loss": 2.2803,
|
| 192 |
+
"step": 25
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.034666666666666665,
|
| 196 |
+
"grad_norm": 0.37337797248642607,
|
| 197 |
+
"learning_rate": 2.9998905094297686e-05,
|
| 198 |
+
"loss": 2.188,
|
| 199 |
+
"step": 26
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.036,
|
| 203 |
+
"grad_norm": 0.16596641795830927,
|
| 204 |
+
"learning_rate": 2.9998509718957563e-05,
|
| 205 |
+
"loss": 2.105,
|
| 206 |
+
"step": 27
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.037333333333333336,
|
| 210 |
+
"grad_norm": 0.3163756174860874,
|
| 211 |
+
"learning_rate": 2.9998053521438427e-05,
|
| 212 |
+
"loss": 2.1819,
|
| 213 |
+
"step": 28
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.03866666666666667,
|
| 217 |
+
"grad_norm": 0.73179742586677,
|
| 218 |
+
"learning_rate": 2.9997536503795834e-05,
|
| 219 |
+
"loss": 2.2281,
|
| 220 |
+
"step": 29
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.04,
|
| 224 |
+
"grad_norm": 0.18793418314598326,
|
| 225 |
+
"learning_rate": 2.9996958668359386e-05,
|
| 226 |
+
"loss": 2.1174,
|
| 227 |
+
"step": 30
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.04133333333333333,
|
| 231 |
+
"grad_norm": 0.6403343407451153,
|
| 232 |
+
"learning_rate": 2.999632001773272e-05,
|
| 233 |
+
"loss": 2.3419,
|
| 234 |
+
"step": 31
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.042666666666666665,
|
| 238 |
+
"grad_norm": 0.2773466547599625,
|
| 239 |
+
"learning_rate": 2.9995620554793495e-05,
|
| 240 |
+
"loss": 1.9779,
|
| 241 |
+
"step": 32
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.044,
|
| 245 |
+
"grad_norm": 0.42992322954286805,
|
| 246 |
+
"learning_rate": 2.999486028269338e-05,
|
| 247 |
+
"loss": 1.891,
|
| 248 |
+
"step": 33
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.04533333333333334,
|
| 252 |
+
"grad_norm": 0.3880877285520559,
|
| 253 |
+
"learning_rate": 2.9994039204858043e-05,
|
| 254 |
+
"loss": 1.9488,
|
| 255 |
+
"step": 34
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.04666666666666667,
|
| 259 |
+
"grad_norm": 0.38995745148913935,
|
| 260 |
+
"learning_rate": 2.999315732498714e-05,
|
| 261 |
+
"loss": 2.5013,
|
| 262 |
+
"step": 35
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.048,
|
| 266 |
+
"grad_norm": 0.4228621926559084,
|
| 267 |
+
"learning_rate": 2.999221464705427e-05,
|
| 268 |
+
"loss": 2.1356,
|
| 269 |
+
"step": 36
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.04933333333333333,
|
| 273 |
+
"grad_norm": 0.4663890399818879,
|
| 274 |
+
"learning_rate": 2.9991211175307006e-05,
|
| 275 |
+
"loss": 2.155,
|
| 276 |
+
"step": 37
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.050666666666666665,
|
| 280 |
+
"grad_norm": 0.5287155786829871,
|
| 281 |
+
"learning_rate": 2.9990146914266826e-05,
|
| 282 |
+
"loss": 1.8496,
|
| 283 |
+
"step": 38
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.052,
|
| 287 |
+
"grad_norm": 0.33925018851926647,
|
| 288 |
+
"learning_rate": 2.9989021868729135e-05,
|
| 289 |
+
"loss": 2.1315,
|
| 290 |
+
"step": 39
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.05333333333333334,
|
| 294 |
+
"grad_norm": 0.16522140027488588,
|
| 295 |
+
"learning_rate": 2.99878360437632e-05,
|
| 296 |
+
"loss": 2.0977,
|
| 297 |
+
"step": 40
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.05466666666666667,
|
| 301 |
+
"grad_norm": 0.6355838754971164,
|
| 302 |
+
"learning_rate": 2.998658944471217e-05,
|
| 303 |
+
"loss": 2.1459,
|
| 304 |
+
"step": 41
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.056,
|
| 308 |
+
"grad_norm": 0.26447590316147407,
|
| 309 |
+
"learning_rate": 2.9985282077193026e-05,
|
| 310 |
+
"loss": 2.0833,
|
| 311 |
+
"step": 42
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.05733333333333333,
|
| 315 |
+
"grad_norm": 0.29258668895913753,
|
| 316 |
+
"learning_rate": 2.9983913947096563e-05,
|
| 317 |
+
"loss": 2.2603,
|
| 318 |
+
"step": 43
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.058666666666666666,
|
| 322 |
+
"grad_norm": 0.2920264557478038,
|
| 323 |
+
"learning_rate": 2.9982485060587357e-05,
|
| 324 |
+
"loss": 1.9828,
|
| 325 |
+
"step": 44
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.06,
|
| 329 |
+
"grad_norm": 0.3860467745966656,
|
| 330 |
+
"learning_rate": 2.9980995424103748e-05,
|
| 331 |
+
"loss": 2.4214,
|
| 332 |
+
"step": 45
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.06133333333333333,
|
| 336 |
+
"grad_norm": 0.5757512413804123,
|
| 337 |
+
"learning_rate": 2.9979445044357814e-05,
|
| 338 |
+
"loss": 2.212,
|
| 339 |
+
"step": 46
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.06266666666666666,
|
| 343 |
+
"grad_norm": 0.21304621427134704,
|
| 344 |
+
"learning_rate": 2.9977833928335316e-05,
|
| 345 |
+
"loss": 1.9271,
|
| 346 |
+
"step": 47
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.064,
|
| 350 |
+
"grad_norm": 0.3301257012003079,
|
| 351 |
+
"learning_rate": 2.9976162083295694e-05,
|
| 352 |
+
"loss": 2.0472,
|
| 353 |
+
"step": 48
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.06533333333333333,
|
| 357 |
+
"grad_norm": 0.3008211448973632,
|
| 358 |
+
"learning_rate": 2.9974429516772018e-05,
|
| 359 |
+
"loss": 2.4323,
|
| 360 |
+
"step": 49
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.06666666666666667,
|
| 364 |
+
"grad_norm": 0.26704445986476444,
|
| 365 |
+
"learning_rate": 2.997263623657097e-05,
|
| 366 |
+
"loss": 2.003,
|
| 367 |
+
"step": 50
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.068,
|
| 371 |
+
"grad_norm": 0.24343367493238866,
|
| 372 |
+
"learning_rate": 2.9970782250772786e-05,
|
| 373 |
+
"loss": 1.9911,
|
| 374 |
+
"step": 51
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.06933333333333333,
|
| 378 |
+
"grad_norm": 0.36868774906810614,
|
| 379 |
+
"learning_rate": 2.9968867567731233e-05,
|
| 380 |
+
"loss": 2.3312,
|
| 381 |
+
"step": 52
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.07066666666666667,
|
| 385 |
+
"grad_norm": 0.4346673091737939,
|
| 386 |
+
"learning_rate": 2.9966892196073583e-05,
|
| 387 |
+
"loss": 2.0461,
|
| 388 |
+
"step": 53
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.072,
|
| 392 |
+
"grad_norm": 0.27732062938320917,
|
| 393 |
+
"learning_rate": 2.996485614470054e-05,
|
| 394 |
+
"loss": 1.8782,
|
| 395 |
+
"step": 54
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.07333333333333333,
|
| 399 |
+
"grad_norm": 0.5044098714622329,
|
| 400 |
+
"learning_rate": 2.9962759422786248e-05,
|
| 401 |
+
"loss": 2.3024,
|
| 402 |
+
"step": 55
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.07466666666666667,
|
| 406 |
+
"grad_norm": 0.35584299953539106,
|
| 407 |
+
"learning_rate": 2.9960602039778196e-05,
|
| 408 |
+
"loss": 1.9019,
|
| 409 |
+
"step": 56
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.076,
|
| 413 |
+
"grad_norm": 0.25515628575898364,
|
| 414 |
+
"learning_rate": 2.995838400539723e-05,
|
| 415 |
+
"loss": 2.1745,
|
| 416 |
+
"step": 57
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.07733333333333334,
|
| 420 |
+
"grad_norm": 0.22266755698926835,
|
| 421 |
+
"learning_rate": 2.9956105329637454e-05,
|
| 422 |
+
"loss": 2.3572,
|
| 423 |
+
"step": 58
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.07866666666666666,
|
| 427 |
+
"grad_norm": 0.23590376658875728,
|
| 428 |
+
"learning_rate": 2.9953766022766228e-05,
|
| 429 |
+
"loss": 2.2639,
|
| 430 |
+
"step": 59
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.08,
|
| 434 |
+
"grad_norm": 0.19958974114647998,
|
| 435 |
+
"learning_rate": 2.9951366095324108e-05,
|
| 436 |
+
"loss": 2.2983,
|
| 437 |
+
"step": 60
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.08133333333333333,
|
| 441 |
+
"grad_norm": 0.26170668739549613,
|
| 442 |
+
"learning_rate": 2.994890555812479e-05,
|
| 443 |
+
"loss": 1.7797,
|
| 444 |
+
"step": 61
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.08266666666666667,
|
| 448 |
+
"grad_norm": 0.1947500638232896,
|
| 449 |
+
"learning_rate": 2.9946384422255074e-05,
|
| 450 |
+
"loss": 2.232,
|
| 451 |
+
"step": 62
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.084,
|
| 455 |
+
"grad_norm": 0.18048707660964908,
|
| 456 |
+
"learning_rate": 2.9943802699074796e-05,
|
| 457 |
+
"loss": 2.3487,
|
| 458 |
+
"step": 63
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.08533333333333333,
|
| 462 |
+
"grad_norm": 0.21913225188044025,
|
| 463 |
+
"learning_rate": 2.994116040021681e-05,
|
| 464 |
+
"loss": 2.1964,
|
| 465 |
+
"step": 64
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.08666666666666667,
|
| 469 |
+
"grad_norm": 0.4242826633242974,
|
| 470 |
+
"learning_rate": 2.9938457537586896e-05,
|
| 471 |
+
"loss": 2.3137,
|
| 472 |
+
"step": 65
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.088,
|
| 476 |
+
"grad_norm": 0.2557454392806272,
|
| 477 |
+
"learning_rate": 2.9935694123363727e-05,
|
| 478 |
+
"loss": 2.1458,
|
| 479 |
+
"step": 66
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.08933333333333333,
|
| 483 |
+
"grad_norm": 0.3093808768168103,
|
| 484 |
+
"learning_rate": 2.9932870169998825e-05,
|
| 485 |
+
"loss": 2.1153,
|
| 486 |
+
"step": 67
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.09066666666666667,
|
| 490 |
+
"grad_norm": 0.33819459397450213,
|
| 491 |
+
"learning_rate": 2.9929985690216478e-05,
|
| 492 |
+
"loss": 2.0772,
|
| 493 |
+
"step": 68
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.092,
|
| 497 |
+
"grad_norm": 0.21377958208808176,
|
| 498 |
+
"learning_rate": 2.9927040697013705e-05,
|
| 499 |
+
"loss": 2.1949,
|
| 500 |
+
"step": 69
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.09333333333333334,
|
| 504 |
+
"grad_norm": 0.2652218777029591,
|
| 505 |
+
"learning_rate": 2.9924035203660188e-05,
|
| 506 |
+
"loss": 2.0031,
|
| 507 |
+
"step": 70
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.09466666666666666,
|
| 511 |
+
"grad_norm": 0.23574741210798578,
|
| 512 |
+
"learning_rate": 2.9920969223698202e-05,
|
| 513 |
+
"loss": 1.9449,
|
| 514 |
+
"step": 71
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.096,
|
| 518 |
+
"grad_norm": 0.2226926959097067,
|
| 519 |
+
"learning_rate": 2.991784277094258e-05,
|
| 520 |
+
"loss": 2.031,
|
| 521 |
+
"step": 72
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.09733333333333333,
|
| 525 |
+
"grad_norm": 0.24814250998530846,
|
| 526 |
+
"learning_rate": 2.9914655859480632e-05,
|
| 527 |
+
"loss": 2.0921,
|
| 528 |
+
"step": 73
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.09866666666666667,
|
| 532 |
+
"grad_norm": 0.33914254917033065,
|
| 533 |
+
"learning_rate": 2.991140850367208e-05,
|
| 534 |
+
"loss": 2.2183,
|
| 535 |
+
"step": 74
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.1,
|
| 539 |
+
"grad_norm": 0.25867270370266443,
|
| 540 |
+
"learning_rate": 2.990810071814901e-05,
|
| 541 |
+
"loss": 1.6416,
|
| 542 |
+
"step": 75
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.1,
|
| 546 |
+
"eval_loss": 1.782590389251709,
|
| 547 |
+
"eval_runtime": 98.6583,
|
| 548 |
+
"eval_samples_per_second": 1.014,
|
| 549 |
+
"eval_steps_per_second": 0.253,
|
| 550 |
+
"step": 75
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.10133333333333333,
|
| 554 |
+
"grad_norm": 0.22920181089355457,
|
| 555 |
+
"learning_rate": 2.990473251781578e-05,
|
| 556 |
+
"loss": 2.1863,
|
| 557 |
+
"step": 76
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 0.10266666666666667,
|
| 561 |
+
"grad_norm": 0.18274633553603156,
|
| 562 |
+
"learning_rate": 2.9901303917848977e-05,
|
| 563 |
+
"loss": 2.2399,
|
| 564 |
+
"step": 77
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.104,
|
| 568 |
+
"grad_norm": 0.18220877055008589,
|
| 569 |
+
"learning_rate": 2.9897814933697335e-05,
|
| 570 |
+
"loss": 1.9149,
|
| 571 |
+
"step": 78
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 0.10533333333333333,
|
| 575 |
+
"grad_norm": 0.2021601886002538,
|
| 576 |
+
"learning_rate": 2.9894265581081682e-05,
|
| 577 |
+
"loss": 2.199,
|
| 578 |
+
"step": 79
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.10666666666666667,
|
| 582 |
+
"grad_norm": 3.7616720925364286,
|
| 583 |
+
"learning_rate": 2.989065587599484e-05,
|
| 584 |
+
"loss": 2.1179,
|
| 585 |
+
"step": 80
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.108,
|
| 589 |
+
"grad_norm": 0.23407462358542114,
|
| 590 |
+
"learning_rate": 2.9886985834701577e-05,
|
| 591 |
+
"loss": 2.5661,
|
| 592 |
+
"step": 81
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.10933333333333334,
|
| 596 |
+
"grad_norm": 1.5338568979686038,
|
| 597 |
+
"learning_rate": 2.9883255473738523e-05,
|
| 598 |
+
"loss": 2.2263,
|
| 599 |
+
"step": 82
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.11066666666666666,
|
| 603 |
+
"grad_norm": 0.24500727274505268,
|
| 604 |
+
"learning_rate": 2.9879464809914113e-05,
|
| 605 |
+
"loss": 2.0621,
|
| 606 |
+
"step": 83
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.112,
|
| 610 |
+
"grad_norm": 0.21179059127462146,
|
| 611 |
+
"learning_rate": 2.987561386030848e-05,
|
| 612 |
+
"loss": 2.0356,
|
| 613 |
+
"step": 84
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 0.11333333333333333,
|
| 617 |
+
"grad_norm": 0.3306008087329227,
|
| 618 |
+
"learning_rate": 2.9871702642273404e-05,
|
| 619 |
+
"loss": 2.0848,
|
| 620 |
+
"step": 85
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.11466666666666667,
|
| 624 |
+
"grad_norm": 0.15657087812683942,
|
| 625 |
+
"learning_rate": 2.9867731173432215e-05,
|
| 626 |
+
"loss": 2.2127,
|
| 627 |
+
"step": 86
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.116,
|
| 631 |
+
"grad_norm": 0.1886500437129195,
|
| 632 |
+
"learning_rate": 2.9863699471679743e-05,
|
| 633 |
+
"loss": 1.9854,
|
| 634 |
+
"step": 87
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 0.11733333333333333,
|
| 638 |
+
"grad_norm": 0.22974349537556282,
|
| 639 |
+
"learning_rate": 2.9859607555182206e-05,
|
| 640 |
+
"loss": 2.1972,
|
| 641 |
+
"step": 88
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.11866666666666667,
|
| 645 |
+
"grad_norm": 0.2578002903522265,
|
| 646 |
+
"learning_rate": 2.9855455442377135e-05,
|
| 647 |
+
"loss": 2.1804,
|
| 648 |
+
"step": 89
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.12,
|
| 652 |
+
"grad_norm": 0.24109737247334548,
|
| 653 |
+
"learning_rate": 2.9851243151973314e-05,
|
| 654 |
+
"loss": 1.9523,
|
| 655 |
+
"step": 90
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.12133333333333333,
|
| 659 |
+
"grad_norm": 0.34121465203098633,
|
| 660 |
+
"learning_rate": 2.9846970702950653e-05,
|
| 661 |
+
"loss": 1.9493,
|
| 662 |
+
"step": 91
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.12266666666666666,
|
| 666 |
+
"grad_norm": 0.2069391899522312,
|
| 667 |
+
"learning_rate": 2.9842638114560144e-05,
|
| 668 |
+
"loss": 2.2207,
|
| 669 |
+
"step": 92
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 0.124,
|
| 673 |
+
"grad_norm": 0.41038744334249916,
|
| 674 |
+
"learning_rate": 2.9838245406323763e-05,
|
| 675 |
+
"loss": 2.0954,
|
| 676 |
+
"step": 93
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 0.12533333333333332,
|
| 680 |
+
"grad_norm": 0.2818876029106677,
|
| 681 |
+
"learning_rate": 2.9833792598034362e-05,
|
| 682 |
+
"loss": 2.063,
|
| 683 |
+
"step": 94
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.12666666666666668,
|
| 687 |
+
"grad_norm": 0.3963506275688602,
|
| 688 |
+
"learning_rate": 2.9829279709755597e-05,
|
| 689 |
+
"loss": 1.9915,
|
| 690 |
+
"step": 95
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.128,
|
| 694 |
+
"grad_norm": 0.3357399751778237,
|
| 695 |
+
"learning_rate": 2.9824706761821845e-05,
|
| 696 |
+
"loss": 2.0319,
|
| 697 |
+
"step": 96
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.12933333333333333,
|
| 701 |
+
"grad_norm": 0.2484694376564558,
|
| 702 |
+
"learning_rate": 2.9820073774838092e-05,
|
| 703 |
+
"loss": 1.9593,
|
| 704 |
+
"step": 97
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.13066666666666665,
|
| 708 |
+
"grad_norm": 0.28583010649806945,
|
| 709 |
+
"learning_rate": 2.9815380769679853e-05,
|
| 710 |
+
"loss": 2.0876,
|
| 711 |
+
"step": 98
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.132,
|
| 715 |
+
"grad_norm": 0.2839808516235786,
|
| 716 |
+
"learning_rate": 2.9810627767493083e-05,
|
| 717 |
+
"loss": 2.107,
|
| 718 |
+
"step": 99
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.13333333333333333,
|
| 722 |
+
"grad_norm": 0.22309498057469626,
|
| 723 |
+
"learning_rate": 2.9805814789694065e-05,
|
| 724 |
+
"loss": 2.3163,
|
| 725 |
+
"step": 100
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.13466666666666666,
|
| 729 |
+
"grad_norm": 0.21629704903995522,
|
| 730 |
+
"learning_rate": 2.9800941857969325e-05,
|
| 731 |
+
"loss": 1.9568,
|
| 732 |
+
"step": 101
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.136,
|
| 736 |
+
"grad_norm": 0.2498232144819958,
|
| 737 |
+
"learning_rate": 2.9796008994275533e-05,
|
| 738 |
+
"loss": 2.2436,
|
| 739 |
+
"step": 102
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.13733333333333334,
|
| 743 |
+
"grad_norm": 0.23251794806618262,
|
| 744 |
+
"learning_rate": 2.979101622083941e-05,
|
| 745 |
+
"loss": 2.1077,
|
| 746 |
+
"step": 103
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.13866666666666666,
|
| 750 |
+
"grad_norm": 0.27094215165929786,
|
| 751 |
+
"learning_rate": 2.978596356015761e-05,
|
| 752 |
+
"loss": 2.2429,
|
| 753 |
+
"step": 104
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.14,
|
| 757 |
+
"grad_norm": 0.23381976958310982,
|
| 758 |
+
"learning_rate": 2.978085103499663e-05,
|
| 759 |
+
"loss": 2.1045,
|
| 760 |
+
"step": 105
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.14133333333333334,
|
| 764 |
+
"grad_norm": 0.31137413672950875,
|
| 765 |
+
"learning_rate": 2.9775678668392713e-05,
|
| 766 |
+
"loss": 2.2828,
|
| 767 |
+
"step": 106
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.14266666666666666,
|
| 771 |
+
"grad_norm": 0.2510514954722372,
|
| 772 |
+
"learning_rate": 2.9770446483651735e-05,
|
| 773 |
+
"loss": 2.0488,
|
| 774 |
+
"step": 107
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.144,
|
| 778 |
+
"grad_norm": 0.20677625084575546,
|
| 779 |
+
"learning_rate": 2.976515450434911e-05,
|
| 780 |
+
"loss": 2.0007,
|
| 781 |
+
"step": 108
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.14533333333333334,
|
| 785 |
+
"grad_norm": 0.1441931078695141,
|
| 786 |
+
"learning_rate": 2.9759802754329665e-05,
|
| 787 |
+
"loss": 2.2914,
|
| 788 |
+
"step": 109
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.14666666666666667,
|
| 792 |
+
"grad_norm": 0.2505777324570115,
|
| 793 |
+
"learning_rate": 2.9754391257707555e-05,
|
| 794 |
+
"loss": 2.3766,
|
| 795 |
+
"step": 110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.148,
|
| 799 |
+
"grad_norm": 0.15292718959769924,
|
| 800 |
+
"learning_rate": 2.9748920038866134e-05,
|
| 801 |
+
"loss": 2.0941,
|
| 802 |
+
"step": 111
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.14933333333333335,
|
| 806 |
+
"grad_norm": 0.323283263320876,
|
| 807 |
+
"learning_rate": 2.9743389122457864e-05,
|
| 808 |
+
"loss": 2.249,
|
| 809 |
+
"step": 112
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.15066666666666667,
|
| 813 |
+
"grad_norm": 0.18122528780601804,
|
| 814 |
+
"learning_rate": 2.9737798533404195e-05,
|
| 815 |
+
"loss": 2.225,
|
| 816 |
+
"step": 113
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.152,
|
| 820 |
+
"grad_norm": 0.1959413592260722,
|
| 821 |
+
"learning_rate": 2.9732148296895444e-05,
|
| 822 |
+
"loss": 2.2521,
|
| 823 |
+
"step": 114
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.15333333333333332,
|
| 827 |
+
"grad_norm": 0.3035240193818014,
|
| 828 |
+
"learning_rate": 2.9726438438390702e-05,
|
| 829 |
+
"loss": 1.8693,
|
| 830 |
+
"step": 115
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.15466666666666667,
|
| 834 |
+
"grad_norm": 0.18185372954938864,
|
| 835 |
+
"learning_rate": 2.9720668983617685e-05,
|
| 836 |
+
"loss": 1.7352,
|
| 837 |
+
"step": 116
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.156,
|
| 841 |
+
"grad_norm": 0.22324589503263917,
|
| 842 |
+
"learning_rate": 2.9714839958572674e-05,
|
| 843 |
+
"loss": 2.1327,
|
| 844 |
+
"step": 117
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.15733333333333333,
|
| 848 |
+
"grad_norm": 0.17640470731953065,
|
| 849 |
+
"learning_rate": 2.9708951389520338e-05,
|
| 850 |
+
"loss": 2.3011,
|
| 851 |
+
"step": 118
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.15866666666666668,
|
| 855 |
+
"grad_norm": 0.22566777150432568,
|
| 856 |
+
"learning_rate": 2.970300330299365e-05,
|
| 857 |
+
"loss": 2.2591,
|
| 858 |
+
"step": 119
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.16,
|
| 862 |
+
"grad_norm": 0.26753575405847074,
|
| 863 |
+
"learning_rate": 2.9696995725793764e-05,
|
| 864 |
+
"loss": 1.9764,
|
| 865 |
+
"step": 120
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.16133333333333333,
|
| 869 |
+
"grad_norm": 0.2625838791216609,
|
| 870 |
+
"learning_rate": 2.969092868498988e-05,
|
| 871 |
+
"loss": 2.0906,
|
| 872 |
+
"step": 121
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.16266666666666665,
|
| 876 |
+
"grad_norm": 0.23542440806071088,
|
| 877 |
+
"learning_rate": 2.9684802207919144e-05,
|
| 878 |
+
"loss": 2.331,
|
| 879 |
+
"step": 122
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.164,
|
| 883 |
+
"grad_norm": 0.16614596830245607,
|
| 884 |
+
"learning_rate": 2.9678616322186506e-05,
|
| 885 |
+
"loss": 2.0534,
|
| 886 |
+
"step": 123
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.16533333333333333,
|
| 890 |
+
"grad_norm": 0.2087450409416601,
|
| 891 |
+
"learning_rate": 2.9672371055664598e-05,
|
| 892 |
+
"loss": 2.3241,
|
| 893 |
+
"step": 124
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.16666666666666666,
|
| 897 |
+
"grad_norm": 0.19314135153093137,
|
| 898 |
+
"learning_rate": 2.9666066436493612e-05,
|
| 899 |
+
"loss": 2.0654,
|
| 900 |
+
"step": 125
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.168,
|
| 904 |
+
"grad_norm": 0.21431158762567906,
|
| 905 |
+
"learning_rate": 2.9659702493081184e-05,
|
| 906 |
+
"loss": 1.9973,
|
| 907 |
+
"step": 126
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 0.16933333333333334,
|
| 911 |
+
"grad_norm": 1.9808549742241428,
|
| 912 |
+
"learning_rate": 2.965327925410226e-05,
|
| 913 |
+
"loss": 2.0077,
|
| 914 |
+
"step": 127
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.17066666666666666,
|
| 918 |
+
"grad_norm": 0.26139018103307665,
|
| 919 |
+
"learning_rate": 2.9646796748498934e-05,
|
| 920 |
+
"loss": 2.0224,
|
| 921 |
+
"step": 128
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.172,
|
| 925 |
+
"grad_norm": 0.25005627301032396,
|
| 926 |
+
"learning_rate": 2.9640255005480376e-05,
|
| 927 |
+
"loss": 2.2725,
|
| 928 |
+
"step": 129
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.17333333333333334,
|
| 932 |
+
"grad_norm": 0.15880416331686953,
|
| 933 |
+
"learning_rate": 2.9633654054522655e-05,
|
| 934 |
+
"loss": 2.2907,
|
| 935 |
+
"step": 130
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.17466666666666666,
|
| 939 |
+
"grad_norm": 0.33529634494995547,
|
| 940 |
+
"learning_rate": 2.9626993925368635e-05,
|
| 941 |
+
"loss": 2.0113,
|
| 942 |
+
"step": 131
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.176,
|
| 946 |
+
"grad_norm": 0.28246655756719385,
|
| 947 |
+
"learning_rate": 2.9620274648027805e-05,
|
| 948 |
+
"loss": 1.6602,
|
| 949 |
+
"step": 132
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.17733333333333334,
|
| 953 |
+
"grad_norm": 0.19752740198982274,
|
| 954 |
+
"learning_rate": 2.96134962527762e-05,
|
| 955 |
+
"loss": 2.4049,
|
| 956 |
+
"step": 133
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.17866666666666667,
|
| 960 |
+
"grad_norm": 0.24132626803001633,
|
| 961 |
+
"learning_rate": 2.960665877015619e-05,
|
| 962 |
+
"loss": 1.9882,
|
| 963 |
+
"step": 134
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.18,
|
| 967 |
+
"grad_norm": 0.20387262050547844,
|
| 968 |
+
"learning_rate": 2.959976223097642e-05,
|
| 969 |
+
"loss": 2.1055,
|
| 970 |
+
"step": 135
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.18133333333333335,
|
| 974 |
+
"grad_norm": 0.1624504592181784,
|
| 975 |
+
"learning_rate": 2.9592806666311612e-05,
|
| 976 |
+
"loss": 2.0519,
|
| 977 |
+
"step": 136
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.18266666666666667,
|
| 981 |
+
"grad_norm": 0.17429829196685362,
|
| 982 |
+
"learning_rate": 2.958579210750246e-05,
|
| 983 |
+
"loss": 2.2465,
|
| 984 |
+
"step": 137
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.184,
|
| 988 |
+
"grad_norm": 0.20450521945125338,
|
| 989 |
+
"learning_rate": 2.9578718586155467e-05,
|
| 990 |
+
"loss": 2.0122,
|
| 991 |
+
"step": 138
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.18533333333333332,
|
| 995 |
+
"grad_norm": 0.25553899821370907,
|
| 996 |
+
"learning_rate": 2.9571586134142824e-05,
|
| 997 |
+
"loss": 1.9806,
|
| 998 |
+
"step": 139
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.18666666666666668,
|
| 1002 |
+
"grad_norm": 0.18251828378522442,
|
| 1003 |
+
"learning_rate": 2.956439478360224e-05,
|
| 1004 |
+
"loss": 2.0153,
|
| 1005 |
+
"step": 140
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.188,
|
| 1009 |
+
"grad_norm": 0.27047669211119385,
|
| 1010 |
+
"learning_rate": 2.9557144566936813e-05,
|
| 1011 |
+
"loss": 2.3386,
|
| 1012 |
+
"step": 141
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.18933333333333333,
|
| 1016 |
+
"grad_norm": 0.25054213266869674,
|
| 1017 |
+
"learning_rate": 2.9549835516814905e-05,
|
| 1018 |
+
"loss": 1.8871,
|
| 1019 |
+
"step": 142
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.19066666666666668,
|
| 1023 |
+
"grad_norm": 0.2177646465341283,
|
| 1024 |
+
"learning_rate": 2.9542467666169946e-05,
|
| 1025 |
+
"loss": 1.9861,
|
| 1026 |
+
"step": 143
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.192,
|
| 1030 |
+
"grad_norm": 0.23131873734472552,
|
| 1031 |
+
"learning_rate": 2.953504104820032e-05,
|
| 1032 |
+
"loss": 2.2903,
|
| 1033 |
+
"step": 144
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.19333333333333333,
|
| 1037 |
+
"grad_norm": 0.19112449587530203,
|
| 1038 |
+
"learning_rate": 2.9527555696369217e-05,
|
| 1039 |
+
"loss": 2.079,
|
| 1040 |
+
"step": 145
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.19466666666666665,
|
| 1044 |
+
"grad_norm": 0.25269202301867255,
|
| 1045 |
+
"learning_rate": 2.9520011644404457e-05,
|
| 1046 |
+
"loss": 2.0033,
|
| 1047 |
+
"step": 146
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.196,
|
| 1051 |
+
"grad_norm": 0.27870777753889076,
|
| 1052 |
+
"learning_rate": 2.9512408926298362e-05,
|
| 1053 |
+
"loss": 2.0747,
|
| 1054 |
+
"step": 147
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.19733333333333333,
|
| 1058 |
+
"grad_norm": 0.2356849572111222,
|
| 1059 |
+
"learning_rate": 2.9504747576307594e-05,
|
| 1060 |
+
"loss": 1.9859,
|
| 1061 |
+
"step": 148
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.19866666666666666,
|
| 1065 |
+
"grad_norm": 0.24802404030831268,
|
| 1066 |
+
"learning_rate": 2.9497027628953e-05,
|
| 1067 |
+
"loss": 1.9636,
|
| 1068 |
+
"step": 149
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.2,
|
| 1072 |
+
"grad_norm": 0.2705817304080749,
|
| 1073 |
+
"learning_rate": 2.9489249119019465e-05,
|
| 1074 |
+
"loss": 2.3547,
|
| 1075 |
+
"step": 150
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.2,
|
| 1079 |
+
"eval_loss": 1.764258623123169,
|
| 1080 |
+
"eval_runtime": 98.8204,
|
| 1081 |
+
"eval_samples_per_second": 1.012,
|
| 1082 |
+
"eval_steps_per_second": 0.253,
|
| 1083 |
+
"step": 150
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 0.20133333333333334,
|
| 1087 |
+
"grad_norm": 0.2582615198348039,
|
| 1088 |
+
"learning_rate": 2.948141208155574e-05,
|
| 1089 |
+
"loss": 2.0245,
|
| 1090 |
+
"step": 151
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.20266666666666666,
|
| 1094 |
+
"grad_norm": 0.4907789285423073,
|
| 1095 |
+
"learning_rate": 2.9473516551874283e-05,
|
| 1096 |
+
"loss": 2.0994,
|
| 1097 |
+
"step": 152
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.204,
|
| 1101 |
+
"grad_norm": 0.18922034328138926,
|
| 1102 |
+
"learning_rate": 2.946556256555113e-05,
|
| 1103 |
+
"loss": 2.323,
|
| 1104 |
+
"step": 153
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.20533333333333334,
|
| 1108 |
+
"grad_norm": 0.2033901054041812,
|
| 1109 |
+
"learning_rate": 2.94575501584257e-05,
|
| 1110 |
+
"loss": 2.0412,
|
| 1111 |
+
"step": 154
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.20666666666666667,
|
| 1115 |
+
"grad_norm": 0.44181703724009636,
|
| 1116 |
+
"learning_rate": 2.9449479366600646e-05,
|
| 1117 |
+
"loss": 2.0977,
|
| 1118 |
+
"step": 155
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.208,
|
| 1122 |
+
"grad_norm": 0.23299679001877052,
|
| 1123 |
+
"learning_rate": 2.94413502264417e-05,
|
| 1124 |
+
"loss": 2.1351,
|
| 1125 |
+
"step": 156
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.20933333333333334,
|
| 1129 |
+
"grad_norm": 0.1826511129435199,
|
| 1130 |
+
"learning_rate": 2.94331627745775e-05,
|
| 1131 |
+
"loss": 2.23,
|
| 1132 |
+
"step": 157
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.21066666666666667,
|
| 1136 |
+
"grad_norm": 0.21212768867139842,
|
| 1137 |
+
"learning_rate": 2.9424917047899425e-05,
|
| 1138 |
+
"loss": 2.1412,
|
| 1139 |
+
"step": 158
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 0.212,
|
| 1143 |
+
"grad_norm": 0.3826582810934168,
|
| 1144 |
+
"learning_rate": 2.9416613083561428e-05,
|
| 1145 |
+
"loss": 2.2128,
|
| 1146 |
+
"step": 159
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 0.21333333333333335,
|
| 1150 |
+
"grad_norm": 0.19962942232450726,
|
| 1151 |
+
"learning_rate": 2.9408250918979886e-05,
|
| 1152 |
+
"loss": 1.9778,
|
| 1153 |
+
"step": 160
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.21466666666666667,
|
| 1157 |
+
"grad_norm": 0.1694294631609089,
|
| 1158 |
+
"learning_rate": 2.9399830591833407e-05,
|
| 1159 |
+
"loss": 2.4602,
|
| 1160 |
+
"step": 161
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.216,
|
| 1164 |
+
"grad_norm": 0.2899013004381108,
|
| 1165 |
+
"learning_rate": 2.9391352140062668e-05,
|
| 1166 |
+
"loss": 1.9118,
|
| 1167 |
+
"step": 162
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 0.21733333333333332,
|
| 1171 |
+
"grad_norm": 0.23107679488781446,
|
| 1172 |
+
"learning_rate": 2.9382815601870252e-05,
|
| 1173 |
+
"loss": 2.0482,
|
| 1174 |
+
"step": 163
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.21866666666666668,
|
| 1178 |
+
"grad_norm": 0.21890406305674734,
|
| 1179 |
+
"learning_rate": 2.9374221015720465e-05,
|
| 1180 |
+
"loss": 2.041,
|
| 1181 |
+
"step": 164
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 0.22,
|
| 1185 |
+
"grad_norm": 0.23483748269482313,
|
| 1186 |
+
"learning_rate": 2.9365568420339173e-05,
|
| 1187 |
+
"loss": 2.2775,
|
| 1188 |
+
"step": 165
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 0.22133333333333333,
|
| 1192 |
+
"grad_norm": 0.2872971245253849,
|
| 1193 |
+
"learning_rate": 2.9356857854713628e-05,
|
| 1194 |
+
"loss": 1.9568,
|
| 1195 |
+
"step": 166
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 0.22266666666666668,
|
| 1199 |
+
"grad_norm": 0.2372348840011155,
|
| 1200 |
+
"learning_rate": 2.9348089358092266e-05,
|
| 1201 |
+
"loss": 2.262,
|
| 1202 |
+
"step": 167
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.224,
|
| 1206 |
+
"grad_norm": 0.36284791627622115,
|
| 1207 |
+
"learning_rate": 2.9339262969984575e-05,
|
| 1208 |
+
"loss": 2.083,
|
| 1209 |
+
"step": 168
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.22533333333333333,
|
| 1213 |
+
"grad_norm": 0.29336818899957795,
|
| 1214 |
+
"learning_rate": 2.9330378730160882e-05,
|
| 1215 |
+
"loss": 1.9668,
|
| 1216 |
+
"step": 169
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.22666666666666666,
|
| 1220 |
+
"grad_norm": 0.2201337837382957,
|
| 1221 |
+
"learning_rate": 2.932143667865218e-05,
|
| 1222 |
+
"loss": 2.2411,
|
| 1223 |
+
"step": 170
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 0.228,
|
| 1227 |
+
"grad_norm": 0.23312129174527574,
|
| 1228 |
+
"learning_rate": 2.931243685574997e-05,
|
| 1229 |
+
"loss": 1.7557,
|
| 1230 |
+
"step": 171
|
| 1231 |
+
},
|
| 1232 |
+
{
|
| 1233 |
+
"epoch": 0.22933333333333333,
|
| 1234 |
+
"grad_norm": 0.2158943628320117,
|
| 1235 |
+
"learning_rate": 2.930337930200603e-05,
|
| 1236 |
+
"loss": 2.2391,
|
| 1237 |
+
"step": 172
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.23066666666666666,
|
| 1241 |
+
"grad_norm": 0.22961082626415702,
|
| 1242 |
+
"learning_rate": 2.929426405823231e-05,
|
| 1243 |
+
"loss": 1.983,
|
| 1244 |
+
"step": 173
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.232,
|
| 1248 |
+
"grad_norm": 0.1589890947108339,
|
| 1249 |
+
"learning_rate": 2.9285091165500653e-05,
|
| 1250 |
+
"loss": 2.2212,
|
| 1251 |
+
"step": 174
|
| 1252 |
+
},
|
| 1253 |
+
{
|
| 1254 |
+
"epoch": 0.23333333333333334,
|
| 1255 |
+
"grad_norm": 0.3416891731513133,
|
| 1256 |
+
"learning_rate": 2.9275860665142697e-05,
|
| 1257 |
+
"loss": 2.151,
|
| 1258 |
+
"step": 175
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.23466666666666666,
|
| 1262 |
+
"grad_norm": 0.3116792178273567,
|
| 1263 |
+
"learning_rate": 2.9266572598749632e-05,
|
| 1264 |
+
"loss": 1.9779,
|
| 1265 |
+
"step": 176
|
| 1266 |
+
},
|
| 1267 |
+
{
|
| 1268 |
+
"epoch": 0.236,
|
| 1269 |
+
"grad_norm": 0.2507975453340776,
|
| 1270 |
+
"learning_rate": 2.925722700817204e-05,
|
| 1271 |
+
"loss": 1.8828,
|
| 1272 |
+
"step": 177
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.23733333333333334,
|
| 1276 |
+
"grad_norm": 0.2679656189407352,
|
| 1277 |
+
"learning_rate": 2.9247823935519685e-05,
|
| 1278 |
+
"loss": 2.0387,
|
| 1279 |
+
"step": 178
|
| 1280 |
+
},
|
| 1281 |
+
{
|
| 1282 |
+
"epoch": 0.23866666666666667,
|
| 1283 |
+
"grad_norm": 0.23989994031916193,
|
| 1284 |
+
"learning_rate": 2.9238363423161357e-05,
|
| 1285 |
+
"loss": 2.0543,
|
| 1286 |
+
"step": 179
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.24,
|
| 1290 |
+
"grad_norm": 0.1858192959656248,
|
| 1291 |
+
"learning_rate": 2.9228845513724636e-05,
|
| 1292 |
+
"loss": 2.1298,
|
| 1293 |
+
"step": 180
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.24133333333333334,
|
| 1297 |
+
"grad_norm": 0.2634790646951208,
|
| 1298 |
+
"learning_rate": 2.921927025009575e-05,
|
| 1299 |
+
"loss": 1.8913,
|
| 1300 |
+
"step": 181
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 0.24266666666666667,
|
| 1304 |
+
"grad_norm": 0.2022924360136422,
|
| 1305 |
+
"learning_rate": 2.920963767541933e-05,
|
| 1306 |
+
"loss": 2.125,
|
| 1307 |
+
"step": 182
|
| 1308 |
+
},
|
| 1309 |
+
{
|
| 1310 |
+
"epoch": 0.244,
|
| 1311 |
+
"grad_norm": 0.2093674676298739,
|
| 1312 |
+
"learning_rate": 2.919994783309827e-05,
|
| 1313 |
+
"loss": 2.0787,
|
| 1314 |
+
"step": 183
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.24533333333333332,
|
| 1318 |
+
"grad_norm": 0.21965402179914853,
|
| 1319 |
+
"learning_rate": 2.9190200766793476e-05,
|
| 1320 |
+
"loss": 2.2001,
|
| 1321 |
+
"step": 184
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.24666666666666667,
|
| 1325 |
+
"grad_norm": 0.1853707607114335,
|
| 1326 |
+
"learning_rate": 2.9180396520423712e-05,
|
| 1327 |
+
"loss": 2.2625,
|
| 1328 |
+
"step": 185
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.248,
|
| 1332 |
+
"grad_norm": 0.16161978911997635,
|
| 1333 |
+
"learning_rate": 2.9170535138165386e-05,
|
| 1334 |
+
"loss": 2.1841,
|
| 1335 |
+
"step": 186
|
| 1336 |
+
},
|
| 1337 |
+
{
|
| 1338 |
+
"epoch": 0.24933333333333332,
|
| 1339 |
+
"grad_norm": 0.3229805512119213,
|
| 1340 |
+
"learning_rate": 2.9160616664452343e-05,
|
| 1341 |
+
"loss": 2.3706,
|
| 1342 |
+
"step": 187
|
| 1343 |
+
},
|
| 1344 |
+
{
|
| 1345 |
+
"epoch": 0.25066666666666665,
|
| 1346 |
+
"grad_norm": 0.2836753076906196,
|
| 1347 |
+
"learning_rate": 2.915064114397568e-05,
|
| 1348 |
+
"loss": 1.6237,
|
| 1349 |
+
"step": 188
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 0.252,
|
| 1353 |
+
"grad_norm": 0.2698172260057232,
|
| 1354 |
+
"learning_rate": 2.9140608621683537e-05,
|
| 1355 |
+
"loss": 1.6731,
|
| 1356 |
+
"step": 189
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.25333333333333335,
|
| 1360 |
+
"grad_norm": 0.19227100326585553,
|
| 1361 |
+
"learning_rate": 2.913051914278089e-05,
|
| 1362 |
+
"loss": 2.3641,
|
| 1363 |
+
"step": 190
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 0.25466666666666665,
|
| 1367 |
+
"grad_norm": 0.17075536856147475,
|
| 1368 |
+
"learning_rate": 2.9120372752729364e-05,
|
| 1369 |
+
"loss": 2.1405,
|
| 1370 |
+
"step": 191
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.256,
|
| 1374 |
+
"grad_norm": 0.18977706986071732,
|
| 1375 |
+
"learning_rate": 2.9110169497247005e-05,
|
| 1376 |
+
"loss": 2.2104,
|
| 1377 |
+
"step": 192
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 0.25733333333333336,
|
| 1381 |
+
"grad_norm": 0.19137759329665405,
|
| 1382 |
+
"learning_rate": 2.909990942230809e-05,
|
| 1383 |
+
"loss": 2.0225,
|
| 1384 |
+
"step": 193
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 0.25866666666666666,
|
| 1388 |
+
"grad_norm": 1.565607637042905,
|
| 1389 |
+
"learning_rate": 2.9089592574142925e-05,
|
| 1390 |
+
"loss": 1.8544,
|
| 1391 |
+
"step": 194
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 0.26,
|
| 1395 |
+
"grad_norm": 0.19540116729751783,
|
| 1396 |
+
"learning_rate": 2.9079218999237602e-05,
|
| 1397 |
+
"loss": 2.3062,
|
| 1398 |
+
"step": 195
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.2613333333333333,
|
| 1402 |
+
"grad_norm": 0.210417328406617,
|
| 1403 |
+
"learning_rate": 2.9068788744333847e-05,
|
| 1404 |
+
"loss": 2.1931,
|
| 1405 |
+
"step": 196
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 0.26266666666666666,
|
| 1409 |
+
"grad_norm": 0.17528914661443556,
|
| 1410 |
+
"learning_rate": 2.905830185642875e-05,
|
| 1411 |
+
"loss": 2.213,
|
| 1412 |
+
"step": 197
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.264,
|
| 1416 |
+
"grad_norm": 0.19180797982535258,
|
| 1417 |
+
"learning_rate": 2.90477583827746e-05,
|
| 1418 |
+
"loss": 2.0999,
|
| 1419 |
+
"step": 198
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 0.2653333333333333,
|
| 1423 |
+
"grad_norm": 0.20387611359379645,
|
| 1424 |
+
"learning_rate": 2.903715837087864e-05,
|
| 1425 |
+
"loss": 2.4077,
|
| 1426 |
+
"step": 199
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 0.26666666666666666,
|
| 1430 |
+
"grad_norm": 0.37614141580385196,
|
| 1431 |
+
"learning_rate": 2.9026501868502878e-05,
|
| 1432 |
+
"loss": 2.345,
|
| 1433 |
+
"step": 200
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.268,
|
| 1437 |
+
"grad_norm": 0.23854637338124732,
|
| 1438 |
+
"learning_rate": 2.901578892366384e-05,
|
| 1439 |
+
"loss": 1.9616,
|
| 1440 |
+
"step": 201
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.2693333333333333,
|
| 1444 |
+
"grad_norm": 0.360995797147823,
|
| 1445 |
+
"learning_rate": 2.9005019584632385e-05,
|
| 1446 |
+
"loss": 2.2637,
|
| 1447 |
+
"step": 202
|
| 1448 |
+
},
|
| 1449 |
+
{
|
| 1450 |
+
"epoch": 0.27066666666666667,
|
| 1451 |
+
"grad_norm": 0.1964827412506095,
|
| 1452 |
+
"learning_rate": 2.899419389993348e-05,
|
| 1453 |
+
"loss": 2.2153,
|
| 1454 |
+
"step": 203
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.272,
|
| 1458 |
+
"grad_norm": 0.2697880949391648,
|
| 1459 |
+
"learning_rate": 2.8983311918345973e-05,
|
| 1460 |
+
"loss": 1.8934,
|
| 1461 |
+
"step": 204
|
| 1462 |
+
},
|
| 1463 |
+
{
|
| 1464 |
+
"epoch": 0.2733333333333333,
|
| 1465 |
+
"grad_norm": 0.2583392197903811,
|
| 1466 |
+
"learning_rate": 2.8972373688902372e-05,
|
| 1467 |
+
"loss": 2.1552,
|
| 1468 |
+
"step": 205
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"epoch": 0.27466666666666667,
|
| 1472 |
+
"grad_norm": 0.2368677594314249,
|
| 1473 |
+
"learning_rate": 2.8961379260888634e-05,
|
| 1474 |
+
"loss": 2.1976,
|
| 1475 |
+
"step": 206
|
| 1476 |
+
},
|
| 1477 |
+
{
|
| 1478 |
+
"epoch": 0.276,
|
| 1479 |
+
"grad_norm": 0.28279002863897235,
|
| 1480 |
+
"learning_rate": 2.895032868384393e-05,
|
| 1481 |
+
"loss": 2.0461,
|
| 1482 |
+
"step": 207
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.2773333333333333,
|
| 1486 |
+
"grad_norm": 0.17431186704640889,
|
| 1487 |
+
"learning_rate": 2.8939222007560446e-05,
|
| 1488 |
+
"loss": 2.2258,
|
| 1489 |
+
"step": 208
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"epoch": 0.2786666666666667,
|
| 1493 |
+
"grad_norm": 0.2607071149284682,
|
| 1494 |
+
"learning_rate": 2.8928059282083126e-05,
|
| 1495 |
+
"loss": 1.8783,
|
| 1496 |
+
"step": 209
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 0.28,
|
| 1500 |
+
"grad_norm": 0.16929508986851496,
|
| 1501 |
+
"learning_rate": 2.8916840557709474e-05,
|
| 1502 |
+
"loss": 2.21,
|
| 1503 |
+
"step": 210
|
| 1504 |
+
},
|
| 1505 |
+
{
|
| 1506 |
+
"epoch": 0.2813333333333333,
|
| 1507 |
+
"grad_norm": 0.30196081586191537,
|
| 1508 |
+
"learning_rate": 2.8905565884989304e-05,
|
| 1509 |
+
"loss": 2.2047,
|
| 1510 |
+
"step": 211
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 0.2826666666666667,
|
| 1514 |
+
"grad_norm": 0.22938305453707358,
|
| 1515 |
+
"learning_rate": 2.889423531472455e-05,
|
| 1516 |
+
"loss": 2.1314,
|
| 1517 |
+
"step": 212
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.284,
|
| 1521 |
+
"grad_norm": 0.24253705098980116,
|
| 1522 |
+
"learning_rate": 2.8882848897968974e-05,
|
| 1523 |
+
"loss": 1.9166,
|
| 1524 |
+
"step": 213
|
| 1525 |
+
},
|
| 1526 |
+
{
|
| 1527 |
+
"epoch": 0.2853333333333333,
|
| 1528 |
+
"grad_norm": 0.21184383213391744,
|
| 1529 |
+
"learning_rate": 2.8871406686028006e-05,
|
| 1530 |
+
"loss": 2.1445,
|
| 1531 |
+
"step": 214
|
| 1532 |
+
},
|
| 1533 |
+
{
|
| 1534 |
+
"epoch": 0.2866666666666667,
|
| 1535 |
+
"grad_norm": 0.20065002671794552,
|
| 1536 |
+
"learning_rate": 2.885990873045846e-05,
|
| 1537 |
+
"loss": 2.2053,
|
| 1538 |
+
"step": 215
|
| 1539 |
+
},
|
| 1540 |
+
{
|
| 1541 |
+
"epoch": 0.288,
|
| 1542 |
+
"grad_norm": 0.20595618063455937,
|
| 1543 |
+
"learning_rate": 2.884835508306833e-05,
|
| 1544 |
+
"loss": 2.0275,
|
| 1545 |
+
"step": 216
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.28933333333333333,
|
| 1549 |
+
"grad_norm": 0.22663216643758846,
|
| 1550 |
+
"learning_rate": 2.883674579591656e-05,
|
| 1551 |
+
"loss": 2.1399,
|
| 1552 |
+
"step": 217
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 0.2906666666666667,
|
| 1556 |
+
"grad_norm": 0.20459405622917667,
|
| 1557 |
+
"learning_rate": 2.8825080921312775e-05,
|
| 1558 |
+
"loss": 2.0352,
|
| 1559 |
+
"step": 218
|
| 1560 |
+
},
|
| 1561 |
+
{
|
| 1562 |
+
"epoch": 0.292,
|
| 1563 |
+
"grad_norm": 0.24151264290319524,
|
| 1564 |
+
"learning_rate": 2.8813360511817092e-05,
|
| 1565 |
+
"loss": 2.0821,
|
| 1566 |
+
"step": 219
|
| 1567 |
+
},
|
| 1568 |
+
{
|
| 1569 |
+
"epoch": 0.29333333333333333,
|
| 1570 |
+
"grad_norm": 0.2377001051292377,
|
| 1571 |
+
"learning_rate": 2.8801584620239833e-05,
|
| 1572 |
+
"loss": 2.0448,
|
| 1573 |
+
"step": 220
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 0.2946666666666667,
|
| 1577 |
+
"grad_norm": 0.3689003556922912,
|
| 1578 |
+
"learning_rate": 2.878975329964134e-05,
|
| 1579 |
+
"loss": 2.107,
|
| 1580 |
+
"step": 221
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.296,
|
| 1584 |
+
"grad_norm": 0.19280049013174935,
|
| 1585 |
+
"learning_rate": 2.877786660333169e-05,
|
| 1586 |
+
"loss": 2.2571,
|
| 1587 |
+
"step": 222
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.29733333333333334,
|
| 1591 |
+
"grad_norm": 0.25684356289858495,
|
| 1592 |
+
"learning_rate": 2.876592458487049e-05,
|
| 1593 |
+
"loss": 1.9243,
|
| 1594 |
+
"step": 223
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.2986666666666667,
|
| 1598 |
+
"grad_norm": 0.2460391604188184,
|
| 1599 |
+
"learning_rate": 2.8753927298066608e-05,
|
| 1600 |
+
"loss": 2.0996,
|
| 1601 |
+
"step": 224
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.3,
|
| 1605 |
+
"grad_norm": 0.2287761563539284,
|
| 1606 |
+
"learning_rate": 2.8741874796977947e-05,
|
| 1607 |
+
"loss": 1.9114,
|
| 1608 |
+
"step": 225
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.3,
|
| 1612 |
+
"eval_loss": 1.7546015977859497,
|
| 1613 |
+
"eval_runtime": 98.7181,
|
| 1614 |
+
"eval_samples_per_second": 1.013,
|
| 1615 |
+
"eval_steps_per_second": 0.253,
|
| 1616 |
+
"step": 225
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.30133333333333334,
|
| 1620 |
+
"grad_norm": 0.31900423121867244,
|
| 1621 |
+
"learning_rate": 2.8729767135911197e-05,
|
| 1622 |
+
"loss": 1.5513,
|
| 1623 |
+
"step": 226
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 0.30266666666666664,
|
| 1627 |
+
"grad_norm": 0.21099907718331903,
|
| 1628 |
+
"learning_rate": 2.8717604369421587e-05,
|
| 1629 |
+
"loss": 1.8735,
|
| 1630 |
+
"step": 227
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 0.304,
|
| 1634 |
+
"grad_norm": 0.23179334925928516,
|
| 1635 |
+
"learning_rate": 2.8705386552312647e-05,
|
| 1636 |
+
"loss": 2.0279,
|
| 1637 |
+
"step": 228
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 0.30533333333333335,
|
| 1641 |
+
"grad_norm": 0.2249602605454057,
|
| 1642 |
+
"learning_rate": 2.869311373963596e-05,
|
| 1643 |
+
"loss": 2.1506,
|
| 1644 |
+
"step": 229
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 0.30666666666666664,
|
| 1648 |
+
"grad_norm": 0.30614551906788895,
|
| 1649 |
+
"learning_rate": 2.8680785986690903e-05,
|
| 1650 |
+
"loss": 2.0057,
|
| 1651 |
+
"step": 230
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 0.308,
|
| 1655 |
+
"grad_norm": 0.32936171428483624,
|
| 1656 |
+
"learning_rate": 2.86684033490244e-05,
|
| 1657 |
+
"loss": 1.9719,
|
| 1658 |
+
"step": 231
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 0.30933333333333335,
|
| 1662 |
+
"grad_norm": 0.21095264296363714,
|
| 1663 |
+
"learning_rate": 2.8655965882430697e-05,
|
| 1664 |
+
"loss": 2.0526,
|
| 1665 |
+
"step": 232
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 0.31066666666666665,
|
| 1669 |
+
"grad_norm": 0.2506508450731689,
|
| 1670 |
+
"learning_rate": 2.8643473642951066e-05,
|
| 1671 |
+
"loss": 2.034,
|
| 1672 |
+
"step": 233
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 0.312,
|
| 1676 |
+
"grad_norm": 0.16272890488221511,
|
| 1677 |
+
"learning_rate": 2.8630926686873598e-05,
|
| 1678 |
+
"loss": 2.2394,
|
| 1679 |
+
"step": 234
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 0.31333333333333335,
|
| 1683 |
+
"grad_norm": 0.2636216414291072,
|
| 1684 |
+
"learning_rate": 2.8618325070732918e-05,
|
| 1685 |
+
"loss": 1.9559,
|
| 1686 |
+
"step": 235
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 0.31466666666666665,
|
| 1690 |
+
"grad_norm": 0.239263175250845,
|
| 1691 |
+
"learning_rate": 2.860566885130994e-05,
|
| 1692 |
+
"loss": 1.9268,
|
| 1693 |
+
"step": 236
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 0.316,
|
| 1697 |
+
"grad_norm": 0.3034572733475333,
|
| 1698 |
+
"learning_rate": 2.8592958085631616e-05,
|
| 1699 |
+
"loss": 2.4146,
|
| 1700 |
+
"step": 237
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 0.31733333333333336,
|
| 1704 |
+
"grad_norm": 0.22016199356343413,
|
| 1705 |
+
"learning_rate": 2.8580192830970674e-05,
|
| 1706 |
+
"loss": 1.901,
|
| 1707 |
+
"step": 238
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 0.31866666666666665,
|
| 1711 |
+
"grad_norm": 0.23102023640647829,
|
| 1712 |
+
"learning_rate": 2.856737314484536e-05,
|
| 1713 |
+
"loss": 2.1498,
|
| 1714 |
+
"step": 239
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 0.32,
|
| 1718 |
+
"grad_norm": 0.20238380453993177,
|
| 1719 |
+
"learning_rate": 2.8554499085019177e-05,
|
| 1720 |
+
"loss": 2.0954,
|
| 1721 |
+
"step": 240
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 0.32133333333333336,
|
| 1725 |
+
"grad_norm": 0.20808439458283837,
|
| 1726 |
+
"learning_rate": 2.854157070950063e-05,
|
| 1727 |
+
"loss": 2.0783,
|
| 1728 |
+
"step": 241
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 0.32266666666666666,
|
| 1732 |
+
"grad_norm": 0.278146848673199,
|
| 1733 |
+
"learning_rate": 2.8528588076542966e-05,
|
| 1734 |
+
"loss": 1.7518,
|
| 1735 |
+
"step": 242
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 0.324,
|
| 1739 |
+
"grad_norm": 0.20915907784327617,
|
| 1740 |
+
"learning_rate": 2.8515551244643903e-05,
|
| 1741 |
+
"loss": 1.7229,
|
| 1742 |
+
"step": 243
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 0.3253333333333333,
|
| 1746 |
+
"grad_norm": 0.5420477506537028,
|
| 1747 |
+
"learning_rate": 2.850246027254537e-05,
|
| 1748 |
+
"loss": 1.758,
|
| 1749 |
+
"step": 244
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 0.32666666666666666,
|
| 1753 |
+
"grad_norm": 0.2800677412041746,
|
| 1754 |
+
"learning_rate": 2.8489315219233248e-05,
|
| 1755 |
+
"loss": 1.9584,
|
| 1756 |
+
"step": 245
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.328,
|
| 1760 |
+
"grad_norm": 0.22569958386724373,
|
| 1761 |
+
"learning_rate": 2.847611614393709e-05,
|
| 1762 |
+
"loss": 2.1033,
|
| 1763 |
+
"step": 246
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 0.3293333333333333,
|
| 1767 |
+
"grad_norm": 0.28225543933402175,
|
| 1768 |
+
"learning_rate": 2.846286310612988e-05,
|
| 1769 |
+
"loss": 2.2579,
|
| 1770 |
+
"step": 247
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 0.33066666666666666,
|
| 1774 |
+
"grad_norm": 0.25637825153268795,
|
| 1775 |
+
"learning_rate": 2.844955616552773e-05,
|
| 1776 |
+
"loss": 1.9495,
|
| 1777 |
+
"step": 248
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 0.332,
|
| 1781 |
+
"grad_norm": 0.3042322338758206,
|
| 1782 |
+
"learning_rate": 2.8436195382089644e-05,
|
| 1783 |
+
"loss": 2.2247,
|
| 1784 |
+
"step": 249
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 0.3333333333333333,
|
| 1788 |
+
"grad_norm": 0.2110291139150889,
|
| 1789 |
+
"learning_rate": 2.8422780816017227e-05,
|
| 1790 |
+
"loss": 2.2582,
|
| 1791 |
+
"step": 250
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 0.33466666666666667,
|
| 1795 |
+
"grad_norm": 0.24476057607988286,
|
| 1796 |
+
"learning_rate": 2.8409312527754417e-05,
|
| 1797 |
+
"loss": 2.1626,
|
| 1798 |
+
"step": 251
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 0.336,
|
| 1802 |
+
"grad_norm": 0.19859582662498323,
|
| 1803 |
+
"learning_rate": 2.8395790577987225e-05,
|
| 1804 |
+
"loss": 2.0592,
|
| 1805 |
+
"step": 252
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 0.3373333333333333,
|
| 1809 |
+
"grad_norm": 0.2560077569832347,
|
| 1810 |
+
"learning_rate": 2.8382215027643447e-05,
|
| 1811 |
+
"loss": 1.9116,
|
| 1812 |
+
"step": 253
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 0.33866666666666667,
|
| 1816 |
+
"grad_norm": 0.2325816082873428,
|
| 1817 |
+
"learning_rate": 2.836858593789239e-05,
|
| 1818 |
+
"loss": 2.1088,
|
| 1819 |
+
"step": 254
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 0.34,
|
| 1823 |
+
"grad_norm": 0.2633745282570676,
|
| 1824 |
+
"learning_rate": 2.8354903370144613e-05,
|
| 1825 |
+
"loss": 2.3106,
|
| 1826 |
+
"step": 255
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 0.3413333333333333,
|
| 1830 |
+
"grad_norm": 0.25738882949934827,
|
| 1831 |
+
"learning_rate": 2.834116738605162e-05,
|
| 1832 |
+
"loss": 1.9449,
|
| 1833 |
+
"step": 256
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 0.3426666666666667,
|
| 1837 |
+
"grad_norm": 0.19022713562039067,
|
| 1838 |
+
"learning_rate": 2.8327378047505625e-05,
|
| 1839 |
+
"loss": 2.0301,
|
| 1840 |
+
"step": 257
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 0.344,
|
| 1844 |
+
"grad_norm": 0.21459535231309906,
|
| 1845 |
+
"learning_rate": 2.8313535416639232e-05,
|
| 1846 |
+
"loss": 2.3784,
|
| 1847 |
+
"step": 258
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 0.3453333333333333,
|
| 1851 |
+
"grad_norm": 0.2725347155581082,
|
| 1852 |
+
"learning_rate": 2.829963955582518e-05,
|
| 1853 |
+
"loss": 1.9749,
|
| 1854 |
+
"step": 259
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 0.3466666666666667,
|
| 1858 |
+
"grad_norm": 0.23824574091343387,
|
| 1859 |
+
"learning_rate": 2.828569052767604e-05,
|
| 1860 |
+
"loss": 2.1253,
|
| 1861 |
+
"step": 260
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 0.348,
|
| 1865 |
+
"grad_norm": 0.30927788905161807,
|
| 1866 |
+
"learning_rate": 2.8271688395043965e-05,
|
| 1867 |
+
"loss": 1.6379,
|
| 1868 |
+
"step": 261
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 0.34933333333333333,
|
| 1872 |
+
"grad_norm": 0.13916235450047149,
|
| 1873 |
+
"learning_rate": 2.8257633221020382e-05,
|
| 1874 |
+
"loss": 2.0781,
|
| 1875 |
+
"step": 262
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 0.3506666666666667,
|
| 1879 |
+
"grad_norm": 0.2038351351610376,
|
| 1880 |
+
"learning_rate": 2.8243525068935705e-05,
|
| 1881 |
+
"loss": 2.1167,
|
| 1882 |
+
"step": 263
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 0.352,
|
| 1886 |
+
"grad_norm": 0.2282060698474947,
|
| 1887 |
+
"learning_rate": 2.8229364002359074e-05,
|
| 1888 |
+
"loss": 1.9761,
|
| 1889 |
+
"step": 264
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 0.35333333333333333,
|
| 1893 |
+
"grad_norm": 0.19533951565078514,
|
| 1894 |
+
"learning_rate": 2.821515008509804e-05,
|
| 1895 |
+
"loss": 2.3902,
|
| 1896 |
+
"step": 265
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 0.3546666666666667,
|
| 1900 |
+
"grad_norm": 0.22231788208405978,
|
| 1901 |
+
"learning_rate": 2.8200883381198297e-05,
|
| 1902 |
+
"loss": 2.2763,
|
| 1903 |
+
"step": 266
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 0.356,
|
| 1907 |
+
"grad_norm": 0.17343255119399525,
|
| 1908 |
+
"learning_rate": 2.818656395494339e-05,
|
| 1909 |
+
"loss": 2.1964,
|
| 1910 |
+
"step": 267
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 0.35733333333333334,
|
| 1914 |
+
"grad_norm": 0.3190889201723669,
|
| 1915 |
+
"learning_rate": 2.817219187085442e-05,
|
| 1916 |
+
"loss": 1.8898,
|
| 1917 |
+
"step": 268
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 0.3586666666666667,
|
| 1921 |
+
"grad_norm": 0.25221058014306336,
|
| 1922 |
+
"learning_rate": 2.8157767193689753e-05,
|
| 1923 |
+
"loss": 2.2137,
|
| 1924 |
+
"step": 269
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 0.36,
|
| 1928 |
+
"grad_norm": 0.25456852526288787,
|
| 1929 |
+
"learning_rate": 2.8143289988444737e-05,
|
| 1930 |
+
"loss": 2.1055,
|
| 1931 |
+
"step": 270
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 0.36133333333333334,
|
| 1935 |
+
"grad_norm": 0.4387988494124866,
|
| 1936 |
+
"learning_rate": 2.8128760320351403e-05,
|
| 1937 |
+
"loss": 2.0767,
|
| 1938 |
+
"step": 271
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 0.3626666666666667,
|
| 1942 |
+
"grad_norm": 0.27199878538165156,
|
| 1943 |
+
"learning_rate": 2.8114178254878156e-05,
|
| 1944 |
+
"loss": 2.1353,
|
| 1945 |
+
"step": 272
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 0.364,
|
| 1949 |
+
"grad_norm": 0.2650737962386425,
|
| 1950 |
+
"learning_rate": 2.8099543857729525e-05,
|
| 1951 |
+
"loss": 2.037,
|
| 1952 |
+
"step": 273
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 0.36533333333333334,
|
| 1956 |
+
"grad_norm": 0.19178775730316125,
|
| 1957 |
+
"learning_rate": 2.808485719484581e-05,
|
| 1958 |
+
"loss": 2.2227,
|
| 1959 |
+
"step": 274
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 0.36666666666666664,
|
| 1963 |
+
"grad_norm": 0.2495538747593823,
|
| 1964 |
+
"learning_rate": 2.8070118332402827e-05,
|
| 1965 |
+
"loss": 2.1111,
|
| 1966 |
+
"step": 275
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 0.368,
|
| 1970 |
+
"grad_norm": 0.23278524231204606,
|
| 1971 |
+
"learning_rate": 2.8055327336811585e-05,
|
| 1972 |
+
"loss": 2.0269,
|
| 1973 |
+
"step": 276
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 0.36933333333333335,
|
| 1977 |
+
"grad_norm": 0.2909743741626987,
|
| 1978 |
+
"learning_rate": 2.804048427471801e-05,
|
| 1979 |
+
"loss": 1.9845,
|
| 1980 |
+
"step": 277
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 0.37066666666666664,
|
| 1984 |
+
"grad_norm": 0.5654501769424853,
|
| 1985 |
+
"learning_rate": 2.8025589213002624e-05,
|
| 1986 |
+
"loss": 1.9375,
|
| 1987 |
+
"step": 278
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 0.372,
|
| 1991 |
+
"grad_norm": 0.24272194924805965,
|
| 1992 |
+
"learning_rate": 2.8010642218780246e-05,
|
| 1993 |
+
"loss": 1.9151,
|
| 1994 |
+
"step": 279
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 0.37333333333333335,
|
| 1998 |
+
"grad_norm": 0.23044549921686833,
|
| 1999 |
+
"learning_rate": 2.7995643359399703e-05,
|
| 2000 |
+
"loss": 2.1577,
|
| 2001 |
+
"step": 280
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 0.37466666666666665,
|
| 2005 |
+
"grad_norm": 0.2297518184027699,
|
| 2006 |
+
"learning_rate": 2.7980592702443518e-05,
|
| 2007 |
+
"loss": 2.2779,
|
| 2008 |
+
"step": 281
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 0.376,
|
| 2012 |
+
"grad_norm": 0.43013668653840365,
|
| 2013 |
+
"learning_rate": 2.79654903157276e-05,
|
| 2014 |
+
"loss": 1.9419,
|
| 2015 |
+
"step": 282
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 0.37733333333333335,
|
| 2019 |
+
"grad_norm": 0.23835905215528563,
|
| 2020 |
+
"learning_rate": 2.795033626730095e-05,
|
| 2021 |
+
"loss": 2.1743,
|
| 2022 |
+
"step": 283
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 0.37866666666666665,
|
| 2026 |
+
"grad_norm": 0.2813058642182358,
|
| 2027 |
+
"learning_rate": 2.793513062544534e-05,
|
| 2028 |
+
"loss": 2.0095,
|
| 2029 |
+
"step": 284
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 0.38,
|
| 2033 |
+
"grad_norm": 0.34406979517562547,
|
| 2034 |
+
"learning_rate": 2.7919873458675022e-05,
|
| 2035 |
+
"loss": 1.9489,
|
| 2036 |
+
"step": 285
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 0.38133333333333336,
|
| 2040 |
+
"grad_norm": 0.32231785755728004,
|
| 2041 |
+
"learning_rate": 2.790456483573642e-05,
|
| 2042 |
+
"loss": 1.58,
|
| 2043 |
+
"step": 286
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 0.38266666666666665,
|
| 2047 |
+
"grad_norm": 0.19333385414604146,
|
| 2048 |
+
"learning_rate": 2.788920482560779e-05,
|
| 2049 |
+
"loss": 1.9752,
|
| 2050 |
+
"step": 287
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 0.384,
|
| 2054 |
+
"grad_norm": 0.1765295794157884,
|
| 2055 |
+
"learning_rate": 2.7873793497498945e-05,
|
| 2056 |
+
"loss": 2.0614,
|
| 2057 |
+
"step": 288
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 0.38533333333333336,
|
| 2061 |
+
"grad_norm": 0.16114993710297087,
|
| 2062 |
+
"learning_rate": 2.7858330920850923e-05,
|
| 2063 |
+
"loss": 2.2809,
|
| 2064 |
+
"step": 289
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 0.38666666666666666,
|
| 2068 |
+
"grad_norm": 0.23825781911565472,
|
| 2069 |
+
"learning_rate": 2.784281716533568e-05,
|
| 2070 |
+
"loss": 2.1522,
|
| 2071 |
+
"step": 290
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 0.388,
|
| 2075 |
+
"grad_norm": 0.17001985217323573,
|
| 2076 |
+
"learning_rate": 2.782725230085579e-05,
|
| 2077 |
+
"loss": 1.9506,
|
| 2078 |
+
"step": 291
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 0.3893333333333333,
|
| 2082 |
+
"grad_norm": 0.26309455211904453,
|
| 2083 |
+
"learning_rate": 2.7811636397544094e-05,
|
| 2084 |
+
"loss": 2.2083,
|
| 2085 |
+
"step": 292
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 0.39066666666666666,
|
| 2089 |
+
"grad_norm": 0.20730249929125907,
|
| 2090 |
+
"learning_rate": 2.7795969525763418e-05,
|
| 2091 |
+
"loss": 2.2347,
|
| 2092 |
+
"step": 293
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 0.392,
|
| 2096 |
+
"grad_norm": 0.24136453807928746,
|
| 2097 |
+
"learning_rate": 2.7780251756106242e-05,
|
| 2098 |
+
"loss": 2.029,
|
| 2099 |
+
"step": 294
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 0.3933333333333333,
|
| 2103 |
+
"grad_norm": 0.35339039676185235,
|
| 2104 |
+
"learning_rate": 2.7764483159394384e-05,
|
| 2105 |
+
"loss": 2.211,
|
| 2106 |
+
"step": 295
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 0.39466666666666667,
|
| 2110 |
+
"grad_norm": 0.25636488066657487,
|
| 2111 |
+
"learning_rate": 2.7748663806678684e-05,
|
| 2112 |
+
"loss": 2.1863,
|
| 2113 |
+
"step": 296
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 0.396,
|
| 2117 |
+
"grad_norm": 0.20584179603071404,
|
| 2118 |
+
"learning_rate": 2.7732793769238674e-05,
|
| 2119 |
+
"loss": 2.2463,
|
| 2120 |
+
"step": 297
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 0.3973333333333333,
|
| 2124 |
+
"grad_norm": 0.23027068378292892,
|
| 2125 |
+
"learning_rate": 2.7716873118582266e-05,
|
| 2126 |
+
"loss": 2.2356,
|
| 2127 |
+
"step": 298
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 0.39866666666666667,
|
| 2131 |
+
"grad_norm": 0.192849837559735,
|
| 2132 |
+
"learning_rate": 2.770090192644543e-05,
|
| 2133 |
+
"loss": 2.2441,
|
| 2134 |
+
"step": 299
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"epoch": 0.4,
|
| 2138 |
+
"grad_norm": 0.2685272097028798,
|
| 2139 |
+
"learning_rate": 2.768488026479187e-05,
|
| 2140 |
+
"loss": 2.0004,
|
| 2141 |
+
"step": 300
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 0.4,
|
| 2145 |
+
"eval_loss": 1.7474404573440552,
|
| 2146 |
+
"eval_runtime": 98.8253,
|
| 2147 |
+
"eval_samples_per_second": 1.012,
|
| 2148 |
+
"eval_steps_per_second": 0.253,
|
| 2149 |
+
"step": 300
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 0.4013333333333333,
|
| 2153 |
+
"grad_norm": 0.26520047905722166,
|
| 2154 |
+
"learning_rate": 2.766880820581269e-05,
|
| 2155 |
+
"loss": 2.0768,
|
| 2156 |
+
"step": 301
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 0.4026666666666667,
|
| 2160 |
+
"grad_norm": 0.31076708165257216,
|
| 2161 |
+
"learning_rate": 2.765268582192608e-05,
|
| 2162 |
+
"loss": 2.0636,
|
| 2163 |
+
"step": 302
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 0.404,
|
| 2167 |
+
"grad_norm": 0.25216548132506456,
|
| 2168 |
+
"learning_rate": 2.763651318577699e-05,
|
| 2169 |
+
"loss": 2.0432,
|
| 2170 |
+
"step": 303
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 0.4053333333333333,
|
| 2174 |
+
"grad_norm": 0.2393514712031434,
|
| 2175 |
+
"learning_rate": 2.7620290370236786e-05,
|
| 2176 |
+
"loss": 2.3461,
|
| 2177 |
+
"step": 304
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 0.4066666666666667,
|
| 2181 |
+
"grad_norm": 0.21599798017132937,
|
| 2182 |
+
"learning_rate": 2.7604017448402954e-05,
|
| 2183 |
+
"loss": 2.3087,
|
| 2184 |
+
"step": 305
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 0.408,
|
| 2188 |
+
"grad_norm": 0.3082071954730338,
|
| 2189 |
+
"learning_rate": 2.7587694493598743e-05,
|
| 2190 |
+
"loss": 1.9302,
|
| 2191 |
+
"step": 306
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 0.4093333333333333,
|
| 2195 |
+
"grad_norm": 0.21921446809175332,
|
| 2196 |
+
"learning_rate": 2.7571321579372835e-05,
|
| 2197 |
+
"loss": 2.1354,
|
| 2198 |
+
"step": 307
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 0.4106666666666667,
|
| 2202 |
+
"grad_norm": 0.2141708374659774,
|
| 2203 |
+
"learning_rate": 2.7554898779499025e-05,
|
| 2204 |
+
"loss": 2.0915,
|
| 2205 |
+
"step": 308
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 0.412,
|
| 2209 |
+
"grad_norm": 0.2615730026982423,
|
| 2210 |
+
"learning_rate": 2.7538426167975895e-05,
|
| 2211 |
+
"loss": 1.9005,
|
| 2212 |
+
"step": 309
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 0.41333333333333333,
|
| 2216 |
+
"grad_norm": 0.24746987092185013,
|
| 2217 |
+
"learning_rate": 2.7521903819026457e-05,
|
| 2218 |
+
"loss": 2.129,
|
| 2219 |
+
"step": 310
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 0.4146666666666667,
|
| 2223 |
+
"grad_norm": 0.23072670442926277,
|
| 2224 |
+
"learning_rate": 2.7505331807097845e-05,
|
| 2225 |
+
"loss": 2.1061,
|
| 2226 |
+
"step": 311
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 0.416,
|
| 2230 |
+
"grad_norm": 3.0090982638780748,
|
| 2231 |
+
"learning_rate": 2.7488710206860944e-05,
|
| 2232 |
+
"loss": 2.1382,
|
| 2233 |
+
"step": 312
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 0.41733333333333333,
|
| 2237 |
+
"grad_norm": 0.2705418179646343,
|
| 2238 |
+
"learning_rate": 2.7472039093210108e-05,
|
| 2239 |
+
"loss": 2.2219,
|
| 2240 |
+
"step": 313
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 0.4186666666666667,
|
| 2244 |
+
"grad_norm": 0.3567890646341258,
|
| 2245 |
+
"learning_rate": 2.7455318541262768e-05,
|
| 2246 |
+
"loss": 1.6885,
|
| 2247 |
+
"step": 314
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 0.42,
|
| 2251 |
+
"grad_norm": 0.20851488835910675,
|
| 2252 |
+
"learning_rate": 2.743854862635912e-05,
|
| 2253 |
+
"loss": 1.947,
|
| 2254 |
+
"step": 315
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 0.42133333333333334,
|
| 2258 |
+
"grad_norm": 0.24844212488624198,
|
| 2259 |
+
"learning_rate": 2.742172942406179e-05,
|
| 2260 |
+
"loss": 2.1564,
|
| 2261 |
+
"step": 316
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 0.4226666666666667,
|
| 2265 |
+
"grad_norm": 0.304277870351501,
|
| 2266 |
+
"learning_rate": 2.7404861010155477e-05,
|
| 2267 |
+
"loss": 1.8723,
|
| 2268 |
+
"step": 317
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 0.424,
|
| 2272 |
+
"grad_norm": 0.20754447108795832,
|
| 2273 |
+
"learning_rate": 2.7387943460646624e-05,
|
| 2274 |
+
"loss": 2.4758,
|
| 2275 |
+
"step": 318
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 0.42533333333333334,
|
| 2279 |
+
"grad_norm": 0.24883308801860618,
|
| 2280 |
+
"learning_rate": 2.7370976851763068e-05,
|
| 2281 |
+
"loss": 1.6462,
|
| 2282 |
+
"step": 319
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 0.4266666666666667,
|
| 2286 |
+
"grad_norm": 0.23024352142002386,
|
| 2287 |
+
"learning_rate": 2.73539612599537e-05,
|
| 2288 |
+
"loss": 2.1038,
|
| 2289 |
+
"step": 320
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 0.428,
|
| 2293 |
+
"grad_norm": 0.3139739867715439,
|
| 2294 |
+
"learning_rate": 2.7336896761888126e-05,
|
| 2295 |
+
"loss": 1.902,
|
| 2296 |
+
"step": 321
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 0.42933333333333334,
|
| 2300 |
+
"grad_norm": 0.21999868040548332,
|
| 2301 |
+
"learning_rate": 2.7319783434456306e-05,
|
| 2302 |
+
"loss": 2.171,
|
| 2303 |
+
"step": 322
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 0.43066666666666664,
|
| 2307 |
+
"grad_norm": 0.2467001853935637,
|
| 2308 |
+
"learning_rate": 2.7302621354768233e-05,
|
| 2309 |
+
"loss": 1.9828,
|
| 2310 |
+
"step": 323
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 0.432,
|
| 2314 |
+
"grad_norm": 0.2496730220022523,
|
| 2315 |
+
"learning_rate": 2.728541060015356e-05,
|
| 2316 |
+
"loss": 2.3413,
|
| 2317 |
+
"step": 324
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 0.43333333333333335,
|
| 2321 |
+
"grad_norm": 0.4056202922850381,
|
| 2322 |
+
"learning_rate": 2.7268151248161253e-05,
|
| 2323 |
+
"loss": 2.0407,
|
| 2324 |
+
"step": 325
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 0.43466666666666665,
|
| 2328 |
+
"grad_norm": 0.2171839709482431,
|
| 2329 |
+
"learning_rate": 2.7250843376559265e-05,
|
| 2330 |
+
"loss": 2.0089,
|
| 2331 |
+
"step": 326
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 0.436,
|
| 2335 |
+
"grad_norm": 0.26937671179427725,
|
| 2336 |
+
"learning_rate": 2.7233487063334172e-05,
|
| 2337 |
+
"loss": 2.2827,
|
| 2338 |
+
"step": 327
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 0.43733333333333335,
|
| 2342 |
+
"grad_norm": 0.20567971005307945,
|
| 2343 |
+
"learning_rate": 2.7216082386690804e-05,
|
| 2344 |
+
"loss": 2.2386,
|
| 2345 |
+
"step": 328
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 0.43866666666666665,
|
| 2349 |
+
"grad_norm": 0.2027173811893726,
|
| 2350 |
+
"learning_rate": 2.7198629425051917e-05,
|
| 2351 |
+
"loss": 2.3031,
|
| 2352 |
+
"step": 329
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 0.44,
|
| 2356 |
+
"grad_norm": 0.2980323024473477,
|
| 2357 |
+
"learning_rate": 2.7181128257057846e-05,
|
| 2358 |
+
"loss": 2.1883,
|
| 2359 |
+
"step": 330
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 0.44133333333333336,
|
| 2363 |
+
"grad_norm": 0.20192851290457803,
|
| 2364 |
+
"learning_rate": 2.716357896156611e-05,
|
| 2365 |
+
"loss": 1.9459,
|
| 2366 |
+
"step": 331
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 0.44266666666666665,
|
| 2370 |
+
"grad_norm": 0.2871519578556366,
|
| 2371 |
+
"learning_rate": 2.7145981617651108e-05,
|
| 2372 |
+
"loss": 2.1742,
|
| 2373 |
+
"step": 332
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 0.444,
|
| 2377 |
+
"grad_norm": 0.1949811627760568,
|
| 2378 |
+
"learning_rate": 2.712833630460372e-05,
|
| 2379 |
+
"loss": 2.2647,
|
| 2380 |
+
"step": 333
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 0.44533333333333336,
|
| 2384 |
+
"grad_norm": 0.2741619826374072,
|
| 2385 |
+
"learning_rate": 2.7110643101930978e-05,
|
| 2386 |
+
"loss": 1.9767,
|
| 2387 |
+
"step": 334
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 0.44666666666666666,
|
| 2391 |
+
"grad_norm": 0.22305808717890685,
|
| 2392 |
+
"learning_rate": 2.7092902089355693e-05,
|
| 2393 |
+
"loss": 1.9271,
|
| 2394 |
+
"step": 335
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 0.448,
|
| 2398 |
+
"grad_norm": 0.31086274795948865,
|
| 2399 |
+
"learning_rate": 2.7075113346816092e-05,
|
| 2400 |
+
"loss": 1.9092,
|
| 2401 |
+
"step": 336
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 0.4493333333333333,
|
| 2405 |
+
"grad_norm": 0.34135122978923976,
|
| 2406 |
+
"learning_rate": 2.7057276954465484e-05,
|
| 2407 |
+
"loss": 1.9028,
|
| 2408 |
+
"step": 337
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 0.45066666666666666,
|
| 2412 |
+
"grad_norm": 0.2422925649756551,
|
| 2413 |
+
"learning_rate": 2.703939299267186e-05,
|
| 2414 |
+
"loss": 2.3585,
|
| 2415 |
+
"step": 338
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 0.452,
|
| 2419 |
+
"grad_norm": 0.21763822152397602,
|
| 2420 |
+
"learning_rate": 2.702146154201757e-05,
|
| 2421 |
+
"loss": 1.9407,
|
| 2422 |
+
"step": 339
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 0.4533333333333333,
|
| 2426 |
+
"grad_norm": 0.26419913958274843,
|
| 2427 |
+
"learning_rate": 2.7003482683298935e-05,
|
| 2428 |
+
"loss": 2.0248,
|
| 2429 |
+
"step": 340
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 0.45466666666666666,
|
| 2433 |
+
"grad_norm": 0.2463717390301776,
|
| 2434 |
+
"learning_rate": 2.698545649752588e-05,
|
| 2435 |
+
"loss": 2.0147,
|
| 2436 |
+
"step": 341
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 0.456,
|
| 2440 |
+
"grad_norm": 0.21124149059138703,
|
| 2441 |
+
"learning_rate": 2.696738306592159e-05,
|
| 2442 |
+
"loss": 1.9508,
|
| 2443 |
+
"step": 342
|
| 2444 |
+
},
|
| 2445 |
+
{
|
| 2446 |
+
"epoch": 0.4573333333333333,
|
| 2447 |
+
"grad_norm": 0.2758662340977994,
|
| 2448 |
+
"learning_rate": 2.694926246992213e-05,
|
| 2449 |
+
"loss": 2.1917,
|
| 2450 |
+
"step": 343
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 0.45866666666666667,
|
| 2454 |
+
"grad_norm": 0.30773115319803324,
|
| 2455 |
+
"learning_rate": 2.693109479117608e-05,
|
| 2456 |
+
"loss": 2.0674,
|
| 2457 |
+
"step": 344
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 0.46,
|
| 2461 |
+
"grad_norm": 0.23761648514026573,
|
| 2462 |
+
"learning_rate": 2.6912880111544163e-05,
|
| 2463 |
+
"loss": 2.0919,
|
| 2464 |
+
"step": 345
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 0.4613333333333333,
|
| 2468 |
+
"grad_norm": 0.25681238620857727,
|
| 2469 |
+
"learning_rate": 2.6894618513098882e-05,
|
| 2470 |
+
"loss": 1.9084,
|
| 2471 |
+
"step": 346
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 0.46266666666666667,
|
| 2475 |
+
"grad_norm": 0.22830005506281845,
|
| 2476 |
+
"learning_rate": 2.687631007812415e-05,
|
| 2477 |
+
"loss": 2.056,
|
| 2478 |
+
"step": 347
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 0.464,
|
| 2482 |
+
"grad_norm": 0.2947866578986673,
|
| 2483 |
+
"learning_rate": 2.6857954889114923e-05,
|
| 2484 |
+
"loss": 1.7343,
|
| 2485 |
+
"step": 348
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 0.4653333333333333,
|
| 2489 |
+
"grad_norm": 0.24624687251309207,
|
| 2490 |
+
"learning_rate": 2.6839553028776817e-05,
|
| 2491 |
+
"loss": 2.2644,
|
| 2492 |
+
"step": 349
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 0.4666666666666667,
|
| 2496 |
+
"grad_norm": 0.23775099815354966,
|
| 2497 |
+
"learning_rate": 2.682110458002575e-05,
|
| 2498 |
+
"loss": 2.0576,
|
| 2499 |
+
"step": 350
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 0.468,
|
| 2503 |
+
"grad_norm": 0.24126580378308465,
|
| 2504 |
+
"learning_rate": 2.6802609625987548e-05,
|
| 2505 |
+
"loss": 2.1331,
|
| 2506 |
+
"step": 351
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 0.4693333333333333,
|
| 2510 |
+
"grad_norm": 0.26152377053961967,
|
| 2511 |
+
"learning_rate": 2.6784068249997586e-05,
|
| 2512 |
+
"loss": 2.194,
|
| 2513 |
+
"step": 352
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 0.4706666666666667,
|
| 2517 |
+
"grad_norm": 0.2670202218764803,
|
| 2518 |
+
"learning_rate": 2.676548053560042e-05,
|
| 2519 |
+
"loss": 1.885,
|
| 2520 |
+
"step": 353
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 0.472,
|
| 2524 |
+
"grad_norm": 0.2227627113395291,
|
| 2525 |
+
"learning_rate": 2.674684656654938e-05,
|
| 2526 |
+
"loss": 2.2919,
|
| 2527 |
+
"step": 354
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 0.47333333333333333,
|
| 2531 |
+
"grad_norm": 0.5386305755161744,
|
| 2532 |
+
"learning_rate": 2.6728166426806237e-05,
|
| 2533 |
+
"loss": 2.1106,
|
| 2534 |
+
"step": 355
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 0.4746666666666667,
|
| 2538 |
+
"grad_norm": 0.2713934931123545,
|
| 2539 |
+
"learning_rate": 2.6709440200540778e-05,
|
| 2540 |
+
"loss": 1.9203,
|
| 2541 |
+
"step": 356
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 0.476,
|
| 2545 |
+
"grad_norm": 0.2687463890320937,
|
| 2546 |
+
"learning_rate": 2.669066797213046e-05,
|
| 2547 |
+
"loss": 2.0278,
|
| 2548 |
+
"step": 357
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 0.47733333333333333,
|
| 2552 |
+
"grad_norm": 0.2420523080214261,
|
| 2553 |
+
"learning_rate": 2.6671849826160018e-05,
|
| 2554 |
+
"loss": 1.9393,
|
| 2555 |
+
"step": 358
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 0.4786666666666667,
|
| 2559 |
+
"grad_norm": 0.2400087971667011,
|
| 2560 |
+
"learning_rate": 2.6652985847421074e-05,
|
| 2561 |
+
"loss": 2.1867,
|
| 2562 |
+
"step": 359
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 0.48,
|
| 2566 |
+
"grad_norm": 0.28320805189191467,
|
| 2567 |
+
"learning_rate": 2.663407612091178e-05,
|
| 2568 |
+
"loss": 1.8562,
|
| 2569 |
+
"step": 360
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 0.48133333333333334,
|
| 2573 |
+
"grad_norm": 0.20819758118379753,
|
| 2574 |
+
"learning_rate": 2.6615120731836412e-05,
|
| 2575 |
+
"loss": 2.0647,
|
| 2576 |
+
"step": 361
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 0.4826666666666667,
|
| 2580 |
+
"grad_norm": 0.19756628127478343,
|
| 2581 |
+
"learning_rate": 2.6596119765604996e-05,
|
| 2582 |
+
"loss": 2.1335,
|
| 2583 |
+
"step": 362
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 0.484,
|
| 2587 |
+
"grad_norm": 0.26928306040248223,
|
| 2588 |
+
"learning_rate": 2.6577073307832925e-05,
|
| 2589 |
+
"loss": 2.0874,
|
| 2590 |
+
"step": 363
|
| 2591 |
+
},
|
| 2592 |
+
{
|
| 2593 |
+
"epoch": 0.48533333333333334,
|
| 2594 |
+
"grad_norm": 0.2164920362315045,
|
| 2595 |
+
"learning_rate": 2.655798144434056e-05,
|
| 2596 |
+
"loss": 2.1714,
|
| 2597 |
+
"step": 364
|
| 2598 |
+
},
|
| 2599 |
+
{
|
| 2600 |
+
"epoch": 0.4866666666666667,
|
| 2601 |
+
"grad_norm": 0.24919806333780237,
|
| 2602 |
+
"learning_rate": 2.6538844261152863e-05,
|
| 2603 |
+
"loss": 1.9509,
|
| 2604 |
+
"step": 365
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 0.488,
|
| 2608 |
+
"grad_norm": 0.27032556044133776,
|
| 2609 |
+
"learning_rate": 2.6519661844498997e-05,
|
| 2610 |
+
"loss": 1.801,
|
| 2611 |
+
"step": 366
|
| 2612 |
+
},
|
| 2613 |
+
{
|
| 2614 |
+
"epoch": 0.48933333333333334,
|
| 2615 |
+
"grad_norm": 0.26804357878114354,
|
| 2616 |
+
"learning_rate": 2.650043428081194e-05,
|
| 2617 |
+
"loss": 2.1762,
|
| 2618 |
+
"step": 367
|
| 2619 |
+
},
|
| 2620 |
+
{
|
| 2621 |
+
"epoch": 0.49066666666666664,
|
| 2622 |
+
"grad_norm": 0.3572135575195324,
|
| 2623 |
+
"learning_rate": 2.6481161656728093e-05,
|
| 2624 |
+
"loss": 1.9907,
|
| 2625 |
+
"step": 368
|
| 2626 |
+
},
|
| 2627 |
+
{
|
| 2628 |
+
"epoch": 0.492,
|
| 2629 |
+
"grad_norm": 0.20724767007565056,
|
| 2630 |
+
"learning_rate": 2.646184405908689e-05,
|
| 2631 |
+
"loss": 2.0845,
|
| 2632 |
+
"step": 369
|
| 2633 |
+
},
|
| 2634 |
+
{
|
| 2635 |
+
"epoch": 0.49333333333333335,
|
| 2636 |
+
"grad_norm": 0.29461252876218635,
|
| 2637 |
+
"learning_rate": 2.6442481574930417e-05,
|
| 2638 |
+
"loss": 2.0182,
|
| 2639 |
+
"step": 370
|
| 2640 |
+
},
|
| 2641 |
+
{
|
| 2642 |
+
"epoch": 0.49466666666666664,
|
| 2643 |
+
"grad_norm": 0.32302958691645034,
|
| 2644 |
+
"learning_rate": 2.6423074291503e-05,
|
| 2645 |
+
"loss": 1.8637,
|
| 2646 |
+
"step": 371
|
| 2647 |
+
},
|
| 2648 |
+
{
|
| 2649 |
+
"epoch": 0.496,
|
| 2650 |
+
"grad_norm": 0.21633487662675288,
|
| 2651 |
+
"learning_rate": 2.6403622296250843e-05,
|
| 2652 |
+
"loss": 2.1181,
|
| 2653 |
+
"step": 372
|
| 2654 |
+
},
|
| 2655 |
+
{
|
| 2656 |
+
"epoch": 0.49733333333333335,
|
| 2657 |
+
"grad_norm": 0.23982202171517725,
|
| 2658 |
+
"learning_rate": 2.6384125676821594e-05,
|
| 2659 |
+
"loss": 2.2663,
|
| 2660 |
+
"step": 373
|
| 2661 |
+
},
|
| 2662 |
+
{
|
| 2663 |
+
"epoch": 0.49866666666666665,
|
| 2664 |
+
"grad_norm": 0.26674040749443334,
|
| 2665 |
+
"learning_rate": 2.636458452106398e-05,
|
| 2666 |
+
"loss": 1.9904,
|
| 2667 |
+
"step": 374
|
| 2668 |
+
},
|
| 2669 |
+
{
|
| 2670 |
+
"epoch": 0.5,
|
| 2671 |
+
"grad_norm": 0.20615794890547487,
|
| 2672 |
+
"learning_rate": 2.63449989170274e-05,
|
| 2673 |
+
"loss": 2.2052,
|
| 2674 |
+
"step": 375
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"epoch": 0.5,
|
| 2678 |
+
"eval_loss": 1.7428412437438965,
|
| 2679 |
+
"eval_runtime": 98.8155,
|
| 2680 |
+
"eval_samples_per_second": 1.012,
|
| 2681 |
+
"eval_steps_per_second": 0.253,
|
| 2682 |
+
"step": 375
|
| 2683 |
+
}
|
| 2684 |
+
],
|
| 2685 |
+
"logging_steps": 1,
|
| 2686 |
+
"max_steps": 1500,
|
| 2687 |
+
"num_input_tokens_seen": 0,
|
| 2688 |
+
"num_train_epochs": 2,
|
| 2689 |
+
"save_steps": 375,
|
| 2690 |
+
"stateful_callbacks": {
|
| 2691 |
+
"TrainerControl": {
|
| 2692 |
+
"args": {
|
| 2693 |
+
"should_epoch_stop": false,
|
| 2694 |
+
"should_evaluate": false,
|
| 2695 |
+
"should_log": false,
|
| 2696 |
+
"should_save": true,
|
| 2697 |
+
"should_training_stop": false
|
| 2698 |
+
},
|
| 2699 |
+
"attributes": {}
|
| 2700 |
+
}
|
| 2701 |
+
},
|
| 2702 |
+
"total_flos": 87275077632000.0,
|
| 2703 |
+
"train_batch_size": 1,
|
| 2704 |
+
"trial_name": null,
|
| 2705 |
+
"trial_params": null
|
| 2706 |
+
}
|
checkpoint-375/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6a971df54a1fea7e372afc85de88dc0b114ebd6f71ee0a1f0c747c7a6c10a7c8
|
| 3 |
+
size 8568
|
checkpoint-375/zero_to_fp32.py
ADDED
|
@@ -0,0 +1,760 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
#!/usr/bin/env python
|
| 2 |
+
|
| 3 |
+
# Copyright (c) Microsoft Corporation.
|
| 4 |
+
# SPDX-License-Identifier: Apache-2.0
|
| 5 |
+
|
| 6 |
+
# DeepSpeed Team
|
| 7 |
+
|
| 8 |
+
# This script extracts fp32 consolidated weights from a zero 1, 2 and 3 DeepSpeed checkpoints. It gets
|
| 9 |
+
# copied into the top level checkpoint dir, so the user can easily do the conversion at any point in
|
| 10 |
+
# the future. Once extracted, the weights don't require DeepSpeed and can be used in any
|
| 11 |
+
# application.
|
| 12 |
+
#
|
| 13 |
+
# example:
|
| 14 |
+
# python zero_to_fp32.py . output_dir/
|
| 15 |
+
# or
|
| 16 |
+
# python zero_to_fp32.py . output_dir/ --safe_serialization
|
| 17 |
+
|
| 18 |
+
import argparse
|
| 19 |
+
import torch
|
| 20 |
+
import glob
|
| 21 |
+
import math
|
| 22 |
+
import os
|
| 23 |
+
import re
|
| 24 |
+
import gc
|
| 25 |
+
import json
|
| 26 |
+
import numpy as np
|
| 27 |
+
from tqdm import tqdm
|
| 28 |
+
from collections import OrderedDict
|
| 29 |
+
from dataclasses import dataclass
|
| 30 |
+
|
| 31 |
+
# while this script doesn't use deepspeed to recover data, since the checkpoints are pickled with
|
| 32 |
+
# DeepSpeed data structures it has to be available in the current python environment.
|
| 33 |
+
from deepspeed.utils import logger
|
| 34 |
+
from deepspeed.checkpoint.constants import (DS_VERSION, OPTIMIZER_STATE_DICT, SINGLE_PARTITION_OF_FP32_GROUPS,
|
| 35 |
+
FP32_FLAT_GROUPS, ZERO_STAGE, PARTITION_COUNT, PARAM_SHAPES, BUFFER_NAMES,
|
| 36 |
+
FROZEN_PARAM_SHAPES, FROZEN_PARAM_FRAGMENTS)
|
| 37 |
+
|
| 38 |
+
|
| 39 |
+
@dataclass
|
| 40 |
+
class zero_model_state:
|
| 41 |
+
buffers: dict()
|
| 42 |
+
param_shapes: dict()
|
| 43 |
+
shared_params: list
|
| 44 |
+
ds_version: int
|
| 45 |
+
frozen_param_shapes: dict()
|
| 46 |
+
frozen_param_fragments: dict()
|
| 47 |
+
|
| 48 |
+
|
| 49 |
+
debug = 0
|
| 50 |
+
|
| 51 |
+
# load to cpu
|
| 52 |
+
device = torch.device('cpu')
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
def atoi(text):
|
| 56 |
+
return int(text) if text.isdigit() else text
|
| 57 |
+
|
| 58 |
+
|
| 59 |
+
def natural_keys(text):
|
| 60 |
+
'''
|
| 61 |
+
alist.sort(key=natural_keys) sorts in human order
|
| 62 |
+
http://nedbatchelder.com/blog/200712/human_sorting.html
|
| 63 |
+
(See Toothy's implementation in the comments)
|
| 64 |
+
'''
|
| 65 |
+
return [atoi(c) for c in re.split(r'(\d+)', text)]
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
def get_model_state_file(checkpoint_dir, zero_stage):
|
| 69 |
+
if not os.path.isdir(checkpoint_dir):
|
| 70 |
+
raise FileNotFoundError(f"Directory '{checkpoint_dir}' doesn't exist")
|
| 71 |
+
|
| 72 |
+
# there should be only one file
|
| 73 |
+
if zero_stage <= 2:
|
| 74 |
+
file = os.path.join(checkpoint_dir, "mp_rank_00_model_states.pt")
|
| 75 |
+
elif zero_stage == 3:
|
| 76 |
+
file = os.path.join(checkpoint_dir, "zero_pp_rank_0_mp_rank_00_model_states.pt")
|
| 77 |
+
|
| 78 |
+
if not os.path.exists(file):
|
| 79 |
+
raise FileNotFoundError(f"can't find model states file at '{file}'")
|
| 80 |
+
|
| 81 |
+
return file
|
| 82 |
+
|
| 83 |
+
|
| 84 |
+
def get_checkpoint_files(checkpoint_dir, glob_pattern):
|
| 85 |
+
# XXX: need to test that this simple glob rule works for multi-node setup too
|
| 86 |
+
ckpt_files = sorted(glob.glob(os.path.join(checkpoint_dir, glob_pattern)), key=natural_keys)
|
| 87 |
+
|
| 88 |
+
if len(ckpt_files) == 0:
|
| 89 |
+
raise FileNotFoundError(f"can't find {glob_pattern} files in directory '{checkpoint_dir}'")
|
| 90 |
+
|
| 91 |
+
return ckpt_files
|
| 92 |
+
|
| 93 |
+
|
| 94 |
+
def get_optim_files(checkpoint_dir):
|
| 95 |
+
return get_checkpoint_files(checkpoint_dir, "*_optim_states.pt")
|
| 96 |
+
|
| 97 |
+
|
| 98 |
+
def get_model_state_files(checkpoint_dir):
|
| 99 |
+
return get_checkpoint_files(checkpoint_dir, "*_model_states.pt")
|
| 100 |
+
|
| 101 |
+
|
| 102 |
+
def parse_model_states(files):
|
| 103 |
+
zero_model_states = []
|
| 104 |
+
for file in files:
|
| 105 |
+
state_dict = torch.load(file, map_location=device, weights_only=False)
|
| 106 |
+
|
| 107 |
+
if BUFFER_NAMES not in state_dict:
|
| 108 |
+
raise ValueError(f"{file} is not a model state checkpoint")
|
| 109 |
+
buffer_names = state_dict[BUFFER_NAMES]
|
| 110 |
+
if debug:
|
| 111 |
+
print("Found buffers:", buffer_names)
|
| 112 |
+
|
| 113 |
+
# recover just the buffers while restoring them to fp32 if they were saved in fp16
|
| 114 |
+
buffers = {k: v.float() for k, v in state_dict["module"].items() if k in buffer_names}
|
| 115 |
+
param_shapes = state_dict[PARAM_SHAPES]
|
| 116 |
+
|
| 117 |
+
# collect parameters that are included in param_shapes
|
| 118 |
+
param_names = []
|
| 119 |
+
for s in param_shapes:
|
| 120 |
+
for name in s.keys():
|
| 121 |
+
param_names.append(name)
|
| 122 |
+
|
| 123 |
+
# update with frozen parameters
|
| 124 |
+
frozen_param_shapes = state_dict.get(FROZEN_PARAM_SHAPES, None)
|
| 125 |
+
if frozen_param_shapes is not None:
|
| 126 |
+
if debug:
|
| 127 |
+
print(f"Found frozen_param_shapes: {frozen_param_shapes}")
|
| 128 |
+
param_names += list(frozen_param_shapes.keys())
|
| 129 |
+
|
| 130 |
+
# handle shared params
|
| 131 |
+
shared_params = [[k, v] for k, v in state_dict["shared_params"].items()]
|
| 132 |
+
|
| 133 |
+
ds_version = state_dict.get(DS_VERSION, None)
|
| 134 |
+
|
| 135 |
+
frozen_param_fragments = state_dict.get(FROZEN_PARAM_FRAGMENTS, None)
|
| 136 |
+
|
| 137 |
+
z_model_state = zero_model_state(buffers=buffers,
|
| 138 |
+
param_shapes=param_shapes,
|
| 139 |
+
shared_params=shared_params,
|
| 140 |
+
ds_version=ds_version,
|
| 141 |
+
frozen_param_shapes=frozen_param_shapes,
|
| 142 |
+
frozen_param_fragments=frozen_param_fragments)
|
| 143 |
+
zero_model_states.append(z_model_state)
|
| 144 |
+
|
| 145 |
+
return zero_model_states
|
| 146 |
+
|
| 147 |
+
|
| 148 |
+
def parse_optim_states(files, ds_checkpoint_dir):
|
| 149 |
+
total_files = len(files)
|
| 150 |
+
state_dicts = []
|
| 151 |
+
for f in tqdm(files, desc='Loading checkpoint shards'):
|
| 152 |
+
state_dict = torch.load(f, map_location=device, mmap=True, weights_only=False)
|
| 153 |
+
# immediately discard the potentially huge 2 optimizer states as we only care for fp32 master weights
|
| 154 |
+
# and also handle the case where it was already removed by another helper script
|
| 155 |
+
state_dict["optimizer_state_dict"].pop("optimizer_state_dict", None)
|
| 156 |
+
state_dicts.append(state_dict)
|
| 157 |
+
|
| 158 |
+
if not ZERO_STAGE in state_dicts[0][OPTIMIZER_STATE_DICT]:
|
| 159 |
+
raise ValueError(f"{files[0]} is not a zero checkpoint")
|
| 160 |
+
zero_stage = state_dicts[0][OPTIMIZER_STATE_DICT][ZERO_STAGE]
|
| 161 |
+
world_size = state_dicts[0][OPTIMIZER_STATE_DICT][PARTITION_COUNT]
|
| 162 |
+
|
| 163 |
+
# For ZeRO-2 each param group can have different partition_count as data parallelism for expert
|
| 164 |
+
# parameters can be different from data parallelism for non-expert parameters. So we can just
|
| 165 |
+
# use the max of the partition_count to get the dp world_size.
|
| 166 |
+
|
| 167 |
+
if type(world_size) is list:
|
| 168 |
+
world_size = max(world_size)
|
| 169 |
+
|
| 170 |
+
if world_size != total_files:
|
| 171 |
+
raise ValueError(
|
| 172 |
+
f"Expected {world_size} of '*_optim_states.pt' under '{ds_checkpoint_dir}' but found {total_files} files. "
|
| 173 |
+
"Possibly due to an overwrite of an old checkpoint, or a checkpoint didn't get saved by one or more processes."
|
| 174 |
+
)
|
| 175 |
+
|
| 176 |
+
# the groups are named differently in each stage
|
| 177 |
+
if zero_stage <= 2:
|
| 178 |
+
fp32_groups_key = SINGLE_PARTITION_OF_FP32_GROUPS
|
| 179 |
+
elif zero_stage == 3:
|
| 180 |
+
fp32_groups_key = FP32_FLAT_GROUPS
|
| 181 |
+
else:
|
| 182 |
+
raise ValueError(f"unknown zero stage {zero_stage}")
|
| 183 |
+
|
| 184 |
+
fp32_flat_groups = [state_dicts[i][OPTIMIZER_STATE_DICT][fp32_groups_key] for i in range(len(state_dicts))]
|
| 185 |
+
return zero_stage, world_size, fp32_flat_groups
|
| 186 |
+
|
| 187 |
+
|
| 188 |
+
def _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters):
|
| 189 |
+
"""
|
| 190 |
+
Returns fp32 state_dict reconstructed from ds checkpoint
|
| 191 |
+
|
| 192 |
+
Args:
|
| 193 |
+
- ``ds_checkpoint_dir``: path to the deepspeed checkpoint folder (where the optimizer files are)
|
| 194 |
+
|
| 195 |
+
"""
|
| 196 |
+
print(f"Processing zero checkpoint '{ds_checkpoint_dir}'")
|
| 197 |
+
|
| 198 |
+
optim_files = get_optim_files(ds_checkpoint_dir)
|
| 199 |
+
zero_stage, world_size, fp32_flat_groups = parse_optim_states(optim_files, ds_checkpoint_dir)
|
| 200 |
+
print(f"Detected checkpoint of type zero stage {zero_stage}, world_size: {world_size}")
|
| 201 |
+
|
| 202 |
+
model_files = get_model_state_files(ds_checkpoint_dir)
|
| 203 |
+
|
| 204 |
+
zero_model_states = parse_model_states(model_files)
|
| 205 |
+
print(f'Parsing checkpoint created by deepspeed=={zero_model_states[0].ds_version}')
|
| 206 |
+
|
| 207 |
+
if zero_stage <= 2:
|
| 208 |
+
return _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 209 |
+
exclude_frozen_parameters)
|
| 210 |
+
elif zero_stage == 3:
|
| 211 |
+
return _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 212 |
+
exclude_frozen_parameters)
|
| 213 |
+
|
| 214 |
+
|
| 215 |
+
def _zero2_merge_frozen_params(state_dict, zero_model_states):
|
| 216 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 217 |
+
return
|
| 218 |
+
|
| 219 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 220 |
+
frozen_param_fragments = zero_model_states[0].frozen_param_fragments
|
| 221 |
+
|
| 222 |
+
if debug:
|
| 223 |
+
num_elem = sum(s.numel() for s in frozen_param_shapes.values())
|
| 224 |
+
print(f'rank 0: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 225 |
+
|
| 226 |
+
wanted_params = len(frozen_param_shapes)
|
| 227 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 228 |
+
avail_numel = sum([p.numel() for p in frozen_param_fragments.values()])
|
| 229 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 230 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 231 |
+
|
| 232 |
+
total_params = 0
|
| 233 |
+
total_numel = 0
|
| 234 |
+
for name, shape in frozen_param_shapes.items():
|
| 235 |
+
total_params += 1
|
| 236 |
+
unpartitioned_numel = shape.numel()
|
| 237 |
+
total_numel += unpartitioned_numel
|
| 238 |
+
|
| 239 |
+
state_dict[name] = frozen_param_fragments[name]
|
| 240 |
+
|
| 241 |
+
if debug:
|
| 242 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 243 |
+
|
| 244 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 245 |
+
|
| 246 |
+
|
| 247 |
+
def _has_callable(obj, fn):
|
| 248 |
+
attr = getattr(obj, fn, None)
|
| 249 |
+
return callable(attr)
|
| 250 |
+
|
| 251 |
+
|
| 252 |
+
def _zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 253 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 254 |
+
|
| 255 |
+
# Reconstruction protocol:
|
| 256 |
+
#
|
| 257 |
+
# XXX: document this
|
| 258 |
+
|
| 259 |
+
if debug:
|
| 260 |
+
for i in range(world_size):
|
| 261 |
+
for j in range(len(fp32_flat_groups[0])):
|
| 262 |
+
print(f"{FP32_FLAT_GROUPS}[{i}][{j}].shape={fp32_flat_groups[i][j].shape}")
|
| 263 |
+
|
| 264 |
+
# XXX: memory usage doubles here (zero2)
|
| 265 |
+
num_param_groups = len(fp32_flat_groups[0])
|
| 266 |
+
merged_single_partition_of_fp32_groups = []
|
| 267 |
+
for i in range(num_param_groups):
|
| 268 |
+
merged_partitions = [sd[i] for sd in fp32_flat_groups]
|
| 269 |
+
full_single_fp32_vector = torch.cat(merged_partitions, 0)
|
| 270 |
+
merged_single_partition_of_fp32_groups.append(full_single_fp32_vector)
|
| 271 |
+
avail_numel = sum(
|
| 272 |
+
[full_single_fp32_vector.numel() for full_single_fp32_vector in merged_single_partition_of_fp32_groups])
|
| 273 |
+
|
| 274 |
+
if debug:
|
| 275 |
+
wanted_params = sum([len(shapes) for shapes in param_shapes])
|
| 276 |
+
wanted_numel = sum([sum(shape.numel() for shape in shapes.values()) for shapes in param_shapes])
|
| 277 |
+
# not asserting if there is a mismatch due to possible padding
|
| 278 |
+
print(f"Have {avail_numel} numels to process.")
|
| 279 |
+
print(f"Need {wanted_numel} numels in {wanted_params} params.")
|
| 280 |
+
|
| 281 |
+
# params
|
| 282 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 283 |
+
# out-of-core computing solution
|
| 284 |
+
total_numel = 0
|
| 285 |
+
total_params = 0
|
| 286 |
+
for shapes, full_single_fp32_vector in zip(param_shapes, merged_single_partition_of_fp32_groups):
|
| 287 |
+
offset = 0
|
| 288 |
+
avail_numel = full_single_fp32_vector.numel()
|
| 289 |
+
for name, shape in shapes.items():
|
| 290 |
+
|
| 291 |
+
unpartitioned_numel = shape.numel() if _has_callable(shape, 'numel') else math.prod(shape)
|
| 292 |
+
total_numel += unpartitioned_numel
|
| 293 |
+
total_params += 1
|
| 294 |
+
|
| 295 |
+
if debug:
|
| 296 |
+
print(f"{name} full shape: {shape} unpartitioned numel {unpartitioned_numel} ")
|
| 297 |
+
state_dict[name] = full_single_fp32_vector.narrow(0, offset, unpartitioned_numel).view(shape)
|
| 298 |
+
offset += unpartitioned_numel
|
| 299 |
+
|
| 300 |
+
# Z2 started to align to 2*world_size to improve nccl performance. Therefore both offset and
|
| 301 |
+
# avail_numel can differ by anywhere between 0..2*world_size. Due to two unrelated complex
|
| 302 |
+
# paddings performed in the code it's almost impossible to predict the exact numbers w/o the
|
| 303 |
+
# live optimizer object, so we are checking that the numbers are within the right range
|
| 304 |
+
align_to = 2 * world_size
|
| 305 |
+
|
| 306 |
+
def zero2_align(x):
|
| 307 |
+
return align_to * math.ceil(x / align_to)
|
| 308 |
+
|
| 309 |
+
if debug:
|
| 310 |
+
print(f"original offset={offset}, avail_numel={avail_numel}")
|
| 311 |
+
|
| 312 |
+
offset = zero2_align(offset)
|
| 313 |
+
avail_numel = zero2_align(avail_numel)
|
| 314 |
+
|
| 315 |
+
if debug:
|
| 316 |
+
print(f"aligned offset={offset}, avail_numel={avail_numel}")
|
| 317 |
+
|
| 318 |
+
# Sanity check
|
| 319 |
+
if offset != avail_numel:
|
| 320 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 321 |
+
|
| 322 |
+
print(f"Reconstructed fp32 state dict with {total_params} params {total_numel} elements")
|
| 323 |
+
|
| 324 |
+
|
| 325 |
+
def _get_fp32_state_dict_from_zero2_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 326 |
+
exclude_frozen_parameters):
|
| 327 |
+
state_dict = OrderedDict()
|
| 328 |
+
|
| 329 |
+
# buffers
|
| 330 |
+
buffers = zero_model_states[0].buffers
|
| 331 |
+
state_dict.update(buffers)
|
| 332 |
+
if debug:
|
| 333 |
+
print(f"added {len(buffers)} buffers")
|
| 334 |
+
|
| 335 |
+
if not exclude_frozen_parameters:
|
| 336 |
+
_zero2_merge_frozen_params(state_dict, zero_model_states)
|
| 337 |
+
|
| 338 |
+
_zero2_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 339 |
+
|
| 340 |
+
# recover shared parameters
|
| 341 |
+
for pair in zero_model_states[0].shared_params:
|
| 342 |
+
if pair[1] in state_dict:
|
| 343 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 344 |
+
|
| 345 |
+
return state_dict
|
| 346 |
+
|
| 347 |
+
|
| 348 |
+
def zero3_partitioned_param_info(unpartitioned_numel, world_size):
|
| 349 |
+
remainder = unpartitioned_numel % world_size
|
| 350 |
+
padding_numel = (world_size - remainder) if remainder else 0
|
| 351 |
+
partitioned_numel = math.ceil(unpartitioned_numel / world_size)
|
| 352 |
+
return partitioned_numel, padding_numel
|
| 353 |
+
|
| 354 |
+
|
| 355 |
+
def _zero3_merge_frozen_params(state_dict, world_size, zero_model_states):
|
| 356 |
+
if zero_model_states[0].frozen_param_shapes is None or len(zero_model_states[0].frozen_param_shapes) == 0:
|
| 357 |
+
return
|
| 358 |
+
|
| 359 |
+
if debug:
|
| 360 |
+
for i in range(world_size):
|
| 361 |
+
num_elem = sum(s.numel() for s in zero_model_states[i].frozen_param_fragments.values())
|
| 362 |
+
print(f'rank {i}: {FROZEN_PARAM_SHAPES}.numel = {num_elem}')
|
| 363 |
+
|
| 364 |
+
frozen_param_shapes = zero_model_states[0].frozen_param_shapes
|
| 365 |
+
wanted_params = len(frozen_param_shapes)
|
| 366 |
+
wanted_numel = sum(s.numel() for s in frozen_param_shapes.values())
|
| 367 |
+
avail_numel = sum([p.numel() for p in zero_model_states[0].frozen_param_fragments.values()]) * world_size
|
| 368 |
+
print(f'Frozen params: Have {avail_numel} numels to process.')
|
| 369 |
+
print(f'Frozen params: Need {wanted_numel} numels in {wanted_params} params')
|
| 370 |
+
|
| 371 |
+
total_params = 0
|
| 372 |
+
total_numel = 0
|
| 373 |
+
for name, shape in zero_model_states[0].frozen_param_shapes.items():
|
| 374 |
+
total_params += 1
|
| 375 |
+
unpartitioned_numel = shape.numel()
|
| 376 |
+
total_numel += unpartitioned_numel
|
| 377 |
+
|
| 378 |
+
param_frags = tuple(model_state.frozen_param_fragments[name] for model_state in zero_model_states)
|
| 379 |
+
state_dict[name] = torch.cat(param_frags, 0).narrow(0, 0, unpartitioned_numel).view(shape)
|
| 380 |
+
|
| 381 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 382 |
+
|
| 383 |
+
if debug:
|
| 384 |
+
print(
|
| 385 |
+
f"Frozen params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 386 |
+
)
|
| 387 |
+
|
| 388 |
+
print(f"Reconstructed Frozen fp32 state dict with {total_params} params {total_numel} elements")
|
| 389 |
+
|
| 390 |
+
|
| 391 |
+
class GatheredTensor:
|
| 392 |
+
"""
|
| 393 |
+
A pseudo tensor that collects partitioned weights.
|
| 394 |
+
It is more memory efficient when there are multiple groups.
|
| 395 |
+
"""
|
| 396 |
+
|
| 397 |
+
def __init__(self, flat_groups, flat_groups_offset, offset, partitioned_numel, shape):
|
| 398 |
+
self.flat_groups = flat_groups
|
| 399 |
+
self.flat_groups_offset = flat_groups_offset
|
| 400 |
+
self.offset = offset
|
| 401 |
+
self.partitioned_numel = partitioned_numel
|
| 402 |
+
self.shape = shape
|
| 403 |
+
self.dtype = self.flat_groups[0][0].dtype
|
| 404 |
+
|
| 405 |
+
def contiguous(self):
|
| 406 |
+
"""
|
| 407 |
+
Merge partitioned weights from flat_groups into a single tensor.
|
| 408 |
+
"""
|
| 409 |
+
end_idx = self.offset + self.partitioned_numel
|
| 410 |
+
world_size = len(self.flat_groups)
|
| 411 |
+
pad_flat_param_chunks = []
|
| 412 |
+
|
| 413 |
+
for rank_i in range(world_size):
|
| 414 |
+
# for each rank, we need to collect weights from related group/groups
|
| 415 |
+
flat_groups_at_rank_i = self.flat_groups[rank_i]
|
| 416 |
+
start_group_id = None
|
| 417 |
+
end_group_id = None
|
| 418 |
+
for group_id in range(len(self.flat_groups_offset)):
|
| 419 |
+
if self.flat_groups_offset[group_id] <= self.offset < self.flat_groups_offset[group_id + 1]:
|
| 420 |
+
start_group_id = group_id
|
| 421 |
+
if self.flat_groups_offset[group_id] < end_idx <= self.flat_groups_offset[group_id + 1]:
|
| 422 |
+
end_group_id = group_id
|
| 423 |
+
break
|
| 424 |
+
# collect weights from related group/groups
|
| 425 |
+
for group_id in range(start_group_id, end_group_id + 1):
|
| 426 |
+
flat_tensor = flat_groups_at_rank_i[group_id]
|
| 427 |
+
start_offset = self.offset - self.flat_groups_offset[group_id]
|
| 428 |
+
end_offset = min(end_idx, self.flat_groups_offset[group_id + 1]) - self.flat_groups_offset[group_id]
|
| 429 |
+
pad_flat_param_chunks.append(flat_tensor[start_offset:end_offset])
|
| 430 |
+
|
| 431 |
+
# collect weights from all ranks
|
| 432 |
+
pad_flat_param = torch.cat(pad_flat_param_chunks, dim=0)
|
| 433 |
+
param = pad_flat_param[:self.shape.numel()].view(self.shape).contiguous()
|
| 434 |
+
return param
|
| 435 |
+
|
| 436 |
+
|
| 437 |
+
def _zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states):
|
| 438 |
+
param_shapes = zero_model_states[0].param_shapes
|
| 439 |
+
avail_numel = sum([flat_group.numel() for flat_group in fp32_flat_groups[0]]) * world_size
|
| 440 |
+
|
| 441 |
+
# Reconstruction protocol: For zero3 we need to zip the partitions together at boundary of each
|
| 442 |
+
# param, re-consolidating each param, while dealing with padding if any
|
| 443 |
+
|
| 444 |
+
# merge list of dicts, preserving order
|
| 445 |
+
param_shapes = {k: v for d in param_shapes for k, v in d.items()}
|
| 446 |
+
|
| 447 |
+
if debug:
|
| 448 |
+
for i in range(world_size):
|
| 449 |
+
print(f"{FP32_FLAT_GROUPS}[{i}].shape={fp32_flat_groups[i].shape}")
|
| 450 |
+
|
| 451 |
+
wanted_params = len(param_shapes)
|
| 452 |
+
wanted_numel = sum(shape.numel() for shape in param_shapes.values())
|
| 453 |
+
# not asserting if there is a mismatch due to possible padding
|
| 454 |
+
avail_numel = fp32_flat_groups[0].numel() * world_size
|
| 455 |
+
print(f"Trainable params: Have {avail_numel} numels to process.")
|
| 456 |
+
print(f"Trainable params: Need {wanted_numel} numels in {wanted_params} params.")
|
| 457 |
+
|
| 458 |
+
# params
|
| 459 |
+
# XXX: for huge models that can't fit into the host's RAM we will have to recode this to support
|
| 460 |
+
# out-of-core computing solution
|
| 461 |
+
offset = 0
|
| 462 |
+
total_numel = 0
|
| 463 |
+
total_params = 0
|
| 464 |
+
flat_groups_offset = [0] + list(np.cumsum([flat_tensor.numel() for flat_tensor in fp32_flat_groups[0]]))
|
| 465 |
+
for name, shape in tqdm(param_shapes.items(), desc='Gathering sharded weights'):
|
| 466 |
+
unpartitioned_numel = shape.numel()
|
| 467 |
+
total_numel += unpartitioned_numel
|
| 468 |
+
total_params += 1
|
| 469 |
+
partitioned_numel, partitioned_padding_numel = zero3_partitioned_param_info(unpartitioned_numel, world_size)
|
| 470 |
+
|
| 471 |
+
if debug:
|
| 472 |
+
print(
|
| 473 |
+
f"Trainable params: {total_params} {name} full shape: {shape} partition0 numel={partitioned_numel} partitioned_padding_numel={partitioned_padding_numel}"
|
| 474 |
+
)
|
| 475 |
+
|
| 476 |
+
# memory efficient tensor
|
| 477 |
+
tensor = GatheredTensor(fp32_flat_groups, flat_groups_offset, offset, partitioned_numel, shape)
|
| 478 |
+
state_dict[name] = tensor
|
| 479 |
+
offset += partitioned_numel
|
| 480 |
+
|
| 481 |
+
offset *= world_size
|
| 482 |
+
|
| 483 |
+
# Sanity check
|
| 484 |
+
if offset != avail_numel:
|
| 485 |
+
raise ValueError(f"consumed {offset} numels out of {avail_numel} - something is wrong")
|
| 486 |
+
|
| 487 |
+
print(f"Reconstructed Trainable fp32 state dict with {total_params} params {total_numel} elements")
|
| 488 |
+
|
| 489 |
+
|
| 490 |
+
def _get_fp32_state_dict_from_zero3_checkpoint(world_size, fp32_flat_groups, zero_model_states,
|
| 491 |
+
exclude_frozen_parameters):
|
| 492 |
+
state_dict = OrderedDict()
|
| 493 |
+
|
| 494 |
+
# buffers
|
| 495 |
+
buffers = zero_model_states[0].buffers
|
| 496 |
+
state_dict.update(buffers)
|
| 497 |
+
if debug:
|
| 498 |
+
print(f"added {len(buffers)} buffers")
|
| 499 |
+
|
| 500 |
+
if not exclude_frozen_parameters:
|
| 501 |
+
_zero3_merge_frozen_params(state_dict, world_size, zero_model_states)
|
| 502 |
+
|
| 503 |
+
_zero3_merge_trainable_params(state_dict, world_size, fp32_flat_groups, zero_model_states)
|
| 504 |
+
|
| 505 |
+
# recover shared parameters
|
| 506 |
+
for pair in zero_model_states[0].shared_params:
|
| 507 |
+
if pair[1] in state_dict:
|
| 508 |
+
state_dict[pair[0]] = state_dict[pair[1]]
|
| 509 |
+
|
| 510 |
+
return state_dict
|
| 511 |
+
|
| 512 |
+
|
| 513 |
+
def to_torch_tensor(state_dict, return_empty_tensor=False):
|
| 514 |
+
"""
|
| 515 |
+
Convert state_dict of GatheredTensor to torch tensor
|
| 516 |
+
"""
|
| 517 |
+
torch_state_dict = {}
|
| 518 |
+
converted_tensors = {}
|
| 519 |
+
for name, tensor in state_dict.items():
|
| 520 |
+
tensor_id = id(tensor)
|
| 521 |
+
if tensor_id in converted_tensors: # shared tensors
|
| 522 |
+
shared_tensor = torch_state_dict[converted_tensors[tensor_id]]
|
| 523 |
+
torch_state_dict[name] = shared_tensor
|
| 524 |
+
else:
|
| 525 |
+
converted_tensors[tensor_id] = name
|
| 526 |
+
if return_empty_tensor:
|
| 527 |
+
torch_state_dict[name] = torch.empty(tensor.shape, dtype=tensor.dtype)
|
| 528 |
+
else:
|
| 529 |
+
torch_state_dict[name] = tensor.contiguous()
|
| 530 |
+
return torch_state_dict
|
| 531 |
+
|
| 532 |
+
|
| 533 |
+
def get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
|
| 534 |
+
tag=None,
|
| 535 |
+
exclude_frozen_parameters=False,
|
| 536 |
+
lazy_mode=False):
|
| 537 |
+
"""
|
| 538 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated state_dict that can be loaded with
|
| 539 |
+
``load_state_dict()`` and used for training without DeepSpeed or shared with others, for example
|
| 540 |
+
via a model hub.
|
| 541 |
+
|
| 542 |
+
Args:
|
| 543 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder
|
| 544 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in 'latest' file. e.g., ``global_step14``
|
| 545 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 546 |
+
- ``lazy_mode``: get state_dict in lazy mode. It returns a dict of pesduo tensor instead of torch tensor, which is more memory efficient.
|
| 547 |
+
Convert the pesduo tensor to torch tensor by ``.contiguous()``
|
| 548 |
+
|
| 549 |
+
Returns:
|
| 550 |
+
- pytorch ``state_dict``
|
| 551 |
+
|
| 552 |
+
A typical usage might be ::
|
| 553 |
+
|
| 554 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 555 |
+
# do the training and checkpoint saving
|
| 556 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir) # already on cpu
|
| 557 |
+
model = model.cpu() # move to cpu
|
| 558 |
+
model.load_state_dict(state_dict)
|
| 559 |
+
# submit to model hub or save the model to share with others
|
| 560 |
+
|
| 561 |
+
In this example the ``model`` will no longer be usable in the deepspeed context of the same
|
| 562 |
+
application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 563 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 564 |
+
|
| 565 |
+
If you want it all done for you, use ``load_state_dict_from_zero_checkpoint`` instead.
|
| 566 |
+
|
| 567 |
+
Note: the above usage may not work if your application doesn't have sufficient free CPU memory.
|
| 568 |
+
You may need to use the offline approach using the ``zero_to_fp32.py`` script that is saved with
|
| 569 |
+
the checkpoint. Or you can load state_dict in lazy mode ::
|
| 570 |
+
|
| 571 |
+
from deepspeed.utils.zero_to_fp32 import get_fp32_state_dict_from_zero_checkpoint
|
| 572 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, lazy_mode=True) # not on cpu
|
| 573 |
+
for name, lazy_tensor in state_dict.item():
|
| 574 |
+
tensor = lazy_tensor.contiguous() # to cpu
|
| 575 |
+
print(name, tensor)
|
| 576 |
+
# del tensor to release memory if it no longer in use
|
| 577 |
+
"""
|
| 578 |
+
if tag is None:
|
| 579 |
+
latest_path = os.path.join(checkpoint_dir, 'latest')
|
| 580 |
+
if os.path.isfile(latest_path):
|
| 581 |
+
with open(latest_path, 'r') as fd:
|
| 582 |
+
tag = fd.read().strip()
|
| 583 |
+
else:
|
| 584 |
+
raise ValueError(f"Unable to find 'latest' file at {latest_path}")
|
| 585 |
+
|
| 586 |
+
ds_checkpoint_dir = os.path.join(checkpoint_dir, tag)
|
| 587 |
+
|
| 588 |
+
if not os.path.isdir(ds_checkpoint_dir):
|
| 589 |
+
raise FileNotFoundError(f"Directory '{ds_checkpoint_dir}' doesn't exist")
|
| 590 |
+
|
| 591 |
+
state_dict = _get_fp32_state_dict_from_zero_checkpoint(ds_checkpoint_dir, exclude_frozen_parameters)
|
| 592 |
+
if lazy_mode:
|
| 593 |
+
return state_dict
|
| 594 |
+
else:
|
| 595 |
+
return to_torch_tensor(state_dict)
|
| 596 |
+
|
| 597 |
+
|
| 598 |
+
def convert_zero_checkpoint_to_fp32_state_dict(checkpoint_dir,
|
| 599 |
+
output_dir,
|
| 600 |
+
max_shard_size="5GB",
|
| 601 |
+
safe_serialization=False,
|
| 602 |
+
tag=None,
|
| 603 |
+
exclude_frozen_parameters=False):
|
| 604 |
+
"""
|
| 605 |
+
Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict`` file that can be
|
| 606 |
+
loaded with ``torch.load(file)`` + ``load_state_dict()`` and used for training without DeepSpeed.
|
| 607 |
+
|
| 608 |
+
Args:
|
| 609 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 610 |
+
- ``output_dir``: directory to the pytorch fp32 state_dict output files
|
| 611 |
+
- ``max_shard_size``: the maximum size for a checkpoint before being sharded, default value is 5GB
|
| 612 |
+
- ``safe_serialization``: whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).
|
| 613 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 614 |
+
- ``exclude_frozen_parameters``: exclude frozen parameters
|
| 615 |
+
"""
|
| 616 |
+
|
| 617 |
+
# Dependency pre-check
|
| 618 |
+
if safe_serialization:
|
| 619 |
+
try:
|
| 620 |
+
from safetensors.torch import save_file
|
| 621 |
+
except ImportError:
|
| 622 |
+
print('If you want to use `safe_serialization`, please `pip install safetensors`')
|
| 623 |
+
raise
|
| 624 |
+
if max_shard_size is not None:
|
| 625 |
+
try:
|
| 626 |
+
from huggingface_hub import split_torch_state_dict_into_shards
|
| 627 |
+
except ImportError:
|
| 628 |
+
print('If you want to use `max_shard_size`, please `pip install huggingface_hub`')
|
| 629 |
+
raise
|
| 630 |
+
|
| 631 |
+
# Convert zero checkpoint to state_dict
|
| 632 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir,
|
| 633 |
+
tag,
|
| 634 |
+
exclude_frozen_parameters,
|
| 635 |
+
lazy_mode=True)
|
| 636 |
+
|
| 637 |
+
# Shard the model if it is too big.
|
| 638 |
+
weights_name = "model.safetensors" if safe_serialization else "pytorch_model.bin"
|
| 639 |
+
if max_shard_size is not None:
|
| 640 |
+
filename_pattern = weights_name.replace(".bin", "{suffix}.bin").replace(".safetensors", "{suffix}.safetensors")
|
| 641 |
+
# an memory-efficient approach for sharding
|
| 642 |
+
empty_state_dict = to_torch_tensor(state_dict, return_empty_tensor=True)
|
| 643 |
+
state_dict_split = split_torch_state_dict_into_shards(empty_state_dict,
|
| 644 |
+
filename_pattern=filename_pattern,
|
| 645 |
+
max_shard_size=max_shard_size)
|
| 646 |
+
else:
|
| 647 |
+
from collections import namedtuple
|
| 648 |
+
StateDictSplit = namedtuple("StateDictSplit", ["is_sharded", "filename_to_tensors"])
|
| 649 |
+
state_dict_split = StateDictSplit(is_sharded=False,
|
| 650 |
+
filename_to_tensors={weights_name: list(state_dict.keys())})
|
| 651 |
+
|
| 652 |
+
# Save the model by shard
|
| 653 |
+
os.makedirs(output_dir, exist_ok=True)
|
| 654 |
+
filename_to_tensors = state_dict_split.filename_to_tensors.items()
|
| 655 |
+
for shard_file, tensors in tqdm(filename_to_tensors, desc="Saving checkpoint shards"):
|
| 656 |
+
shard_state_dict = {tensor_name: state_dict[tensor_name] for tensor_name in tensors}
|
| 657 |
+
shard_state_dict = to_torch_tensor(shard_state_dict)
|
| 658 |
+
output_path = os.path.join(output_dir, shard_file)
|
| 659 |
+
if safe_serialization:
|
| 660 |
+
save_file(shard_state_dict, output_path, metadata={"format": "pt"})
|
| 661 |
+
else:
|
| 662 |
+
torch.save(shard_state_dict, output_path)
|
| 663 |
+
# release the memory of current shard
|
| 664 |
+
for tensor_name in list(shard_state_dict.keys()):
|
| 665 |
+
del state_dict[tensor_name]
|
| 666 |
+
del shard_state_dict[tensor_name]
|
| 667 |
+
del shard_state_dict
|
| 668 |
+
gc.collect()
|
| 669 |
+
|
| 670 |
+
# Save index if sharded
|
| 671 |
+
if state_dict_split.is_sharded:
|
| 672 |
+
index = {
|
| 673 |
+
"metadata": state_dict_split.metadata,
|
| 674 |
+
"weight_map": state_dict_split.tensor_to_filename,
|
| 675 |
+
}
|
| 676 |
+
save_index_file = "model.safetensors.index.json" if safe_serialization else "pytorch_model.bin.index.json"
|
| 677 |
+
save_index_file = os.path.join(output_dir, save_index_file)
|
| 678 |
+
with open(save_index_file, "w", encoding="utf-8") as f:
|
| 679 |
+
content = json.dumps(index, indent=2, sort_keys=True) + "\n"
|
| 680 |
+
f.write(content)
|
| 681 |
+
|
| 682 |
+
|
| 683 |
+
def load_state_dict_from_zero_checkpoint(model, checkpoint_dir, tag=None):
|
| 684 |
+
"""
|
| 685 |
+
1. Put the provided model to cpu
|
| 686 |
+
2. Convert ZeRO 2 or 3 checkpoint into a single fp32 consolidated ``state_dict``
|
| 687 |
+
3. Load it into the provided model
|
| 688 |
+
|
| 689 |
+
Args:
|
| 690 |
+
- ``model``: the model object to update
|
| 691 |
+
- ``checkpoint_dir``: path to the desired checkpoint folder. (one that contains the tag-folder, like ``global_step14``)
|
| 692 |
+
- ``tag``: checkpoint tag used as a unique identifier for checkpoint. If not provided will attempt to load tag in the file named ``latest`` in the checkpoint folder, e.g., ``global_step14``
|
| 693 |
+
|
| 694 |
+
Returns:
|
| 695 |
+
- ``model`: modified model
|
| 696 |
+
|
| 697 |
+
Make sure you have plenty of CPU memory available before you call this function. If you don't
|
| 698 |
+
have enough use the ``zero_to_fp32.py`` utility to do the conversion. You will find it
|
| 699 |
+
conveniently placed for you in the checkpoint folder.
|
| 700 |
+
|
| 701 |
+
A typical usage might be ::
|
| 702 |
+
|
| 703 |
+
from deepspeed.utils.zero_to_fp32 import load_state_dict_from_zero_checkpoint
|
| 704 |
+
model = load_state_dict_from_zero_checkpoint(trainer.model, checkpoint_dir)
|
| 705 |
+
# submit to model hub or save the model to share with others
|
| 706 |
+
|
| 707 |
+
Note, that once this was run, the ``model`` will no longer be usable in the deepspeed context
|
| 708 |
+
of the same application. i.e. you will need to re-initialize the deepspeed engine, since
|
| 709 |
+
``model.load_state_dict(state_dict)`` will remove all the deepspeed magic from it.
|
| 710 |
+
|
| 711 |
+
"""
|
| 712 |
+
logger.info(f"Extracting fp32 weights")
|
| 713 |
+
state_dict = get_fp32_state_dict_from_zero_checkpoint(checkpoint_dir, tag)
|
| 714 |
+
|
| 715 |
+
logger.info(f"Overwriting model with fp32 weights")
|
| 716 |
+
model = model.cpu()
|
| 717 |
+
model.load_state_dict(state_dict, strict=False)
|
| 718 |
+
|
| 719 |
+
return model
|
| 720 |
+
|
| 721 |
+
|
| 722 |
+
if __name__ == "__main__":
|
| 723 |
+
parser = argparse.ArgumentParser()
|
| 724 |
+
parser.add_argument("checkpoint_dir",
|
| 725 |
+
type=str,
|
| 726 |
+
help="path to the desired checkpoint folder, e.g., path/checkpoint-12")
|
| 727 |
+
parser.add_argument("output_dir",
|
| 728 |
+
type=str,
|
| 729 |
+
help="directory to the pytorch fp32 state_dict output files"
|
| 730 |
+
"(e.g. path/checkpoint-12-output/)")
|
| 731 |
+
parser.add_argument(
|
| 732 |
+
"--max_shard_size",
|
| 733 |
+
type=str,
|
| 734 |
+
default="5GB",
|
| 735 |
+
help="The maximum size for a checkpoint before being sharded. Checkpoints shard will then be each of size"
|
| 736 |
+
"lower than this size. If expressed as a string, needs to be digits followed by a unit (like `5MB`"
|
| 737 |
+
"We default it to 5GB in order for models to be able to run easily on free-tier google colab instances"
|
| 738 |
+
"without CPU OOM issues.")
|
| 739 |
+
parser.add_argument(
|
| 740 |
+
"--safe_serialization",
|
| 741 |
+
default=False,
|
| 742 |
+
action='store_true',
|
| 743 |
+
help="Whether to save the model using `safetensors` or the traditional PyTorch way (that uses `pickle`).")
|
| 744 |
+
parser.add_argument("-t",
|
| 745 |
+
"--tag",
|
| 746 |
+
type=str,
|
| 747 |
+
default=None,
|
| 748 |
+
help="checkpoint tag used as a unique identifier for checkpoint. e.g., global_step1")
|
| 749 |
+
parser.add_argument("--exclude_frozen_parameters", action='store_true', help="exclude frozen parameters")
|
| 750 |
+
parser.add_argument("-d", "--debug", action='store_true', help="enable debug")
|
| 751 |
+
args = parser.parse_args()
|
| 752 |
+
|
| 753 |
+
debug = args.debug
|
| 754 |
+
|
| 755 |
+
convert_zero_checkpoint_to_fp32_state_dict(args.checkpoint_dir,
|
| 756 |
+
args.output_dir,
|
| 757 |
+
max_shard_size=args.max_shard_size,
|
| 758 |
+
safe_serialization=args.safe_serialization,
|
| 759 |
+
tag=args.tag,
|
| 760 |
+
exclude_frozen_parameters=args.exclude_frozen_parameters)
|