Total runs:
4.8M
Run Growth:
183.2K
Growth Rate:
3.81%
Sponsored by
PostMcpAi
dphn / dolphin-2.8-mistral-7b-v02
Model's Last Updated: May 20 2024
text-generation
Introduction of dolphin-2.8-mistral-7b-v02
Model Details of dolphin-2.8-mistral-7b-v02
Dolphin 2.8 Mistral 7b v0.2 🐬
By Eric Hartford and Cognitive Computations
My appreciation for the sponsors of Dolphin 2.8:
- Crusoe Cloud - provided excellent on-demand 10xL40S node
- Winston Sou - Along with a generous anonymous sponsor, donated a massive personally owned compute resource!
- Abacus AI - my employer and partner in many things.
This model is based on Mistral-7b-v0.2 a new base model released by MistralAI on March 23, 2024 but they have not yet published on HuggingFace. Thanks to @alpindale for converting / publishing.
The base model has 32k context, and the full-weights fine-tune was with 16k sequence lengths.
It took 3 days on 10x L40S provided by Crusoe Cloud
Dolphin-2.8 has a variety of instruction, conversational, and coding skills.
Dolphin is uncensored. I have filtered the dataset to remove alignment and bias. This makes the model more compliant. You are advised to implement your own alignment layer before exposing the model as a service. It will be highly compliant to any requests, even unethical ones. Please read my blog post about uncensored models. https://erichartford.com/uncensored-models You are responsible for any content you create using this model. Enjoy responsibly.
Dolphin is licensed Apache 2.0. I grant permission for any use including commercial. Dolphin was trained on data generated from GPT4 among other models.
Evals
{
"arc_challenge": {
"acc,none": 0.5921501706484642,
"acc_stderr,none": 0.014361097288449701,
"acc_norm,none": 0.6339590443686007,
"acc_norm_stderr,none": 0.014077223108470139
},
"gsm8k": {
"exact_match,strict-match": 0.4783927217589083,
"exact_match_stderr,strict-match": 0.013759618667051773,
"exact_match,flexible-extract": 0.5367702805155421,
"exact_match_stderr,flexible-extract": 0.013735191956468648
},
"hellaswag": {
"acc,none": 0.6389165504879506,
"acc_stderr,none": 0.004793330525656218,
"acc_norm,none": 0.8338976299541924,
"acc_norm_stderr,none": 0.00371411888431746
},
"mmlu": {
"acc,none": 0.6122347243982339,
"acc_stderr,none": 0.003893774654142997
},
"truthfulqa_mc2": {
"acc,none": 0.5189872652778472,
"acc_stderr,none": 0.014901128316426086
},
"winogrande": {
"acc,none": 0.7971586424625099,
"acc_stderr,none": 0.011301439925936643
}
}
See axolotl config
axolotl version:
0.4.0
base_model: alpindale/Mistral-7B-v0.2-hf
model_type: MistralForCausalLM
tokenizer_type: LlamaTokenizer
is_mistral_derived_model: true
load_in_8bit: false
load_in_4bit: false
strict: false
datasets:
- path: /workspace/datasets/dolphin201-sharegpt2.jsonl
type: sharegpt
- path: /workspace/datasets/dolphin-coder-translate-sharegpt2.jsonl
type: sharegpt
- path: /workspace/datasets/dolphin-coder-codegen-sharegpt2.jsonl
type: sharegpt
- path: /workspace/datasets/m-a-p_Code-Feedback-sharegpt.jsonl
type: sharegpt
- path: /workspace/datasets/m-a-p_CodeFeedback-Filtered-Instruction-sharegpt.jsonl
type: sharegpt
- path: /workspace/datasets/not_samantha_norefusals.jsonl
type: sharegpt
- path: /workspace/datasets/openhermes2_5-sharegpt.jsonl
type: sharegpt
chat_template: chatml
dataset_prepared_path: last_run_prepared
val_set_size: 0.001
output_dir: /workspace/dolphin-2.8-mistral-7b
sequence_len: 16384
sample_packing: true
pad_to_sequence_len: true
wandb_project: dolphin
wandb_entity:
wandb_watch:
wandb_run_id:
wandb_log_model:
gradient_accumulation_steps: 8
micro_batch_size: 3
num_epochs: 4
adam_beta2: 0.95
adam_epsilon: 0.00001
max_grad_norm: 1.0
lr_scheduler: cosine
learning_rate: 0.000005
optimizer: adamw_bnb_8bit
train_on_inputs: false
group_by_length: false
bf16: true
fp16: false
tf32: false
gradient_checkpointing: true
gradient_checkpointing_kwargs:
use_reentrant: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true
warmup_steps: 10
eval_steps: 73
eval_table_size:
eval_table_max_new_tokens:
eval_sample_packing: false
saves_per_epoch:
save_steps: 73
save_total_limit: 2
debug:
deepspeed: deepspeed_configs/zero3_bf16.json
weight_decay: 0.1
fsdp:
fsdp_config:
special_tokens:
eos_token: "<|im_end|>"
tokens:
- "<|im_start|>"
workspace/dolphin-2.8-mistral-7b
This model is a fine-tuned version of alpindale/Mistral-7B-v0.2-hf on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4828
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-06
- train_batch_size: 3
- eval_batch_size: 3
- seed: 42
- distributed_type: multi-GPU
- num_devices: 10
- gradient_accumulation_steps: 8
- total_train_batch_size: 240
- total_eval_batch_size: 30
- optimizer: Adam with betas=(0.9,0.95) and epsilon=1e-05
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 10
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 1.1736 | 0.0 | 1 | 1.0338 |
| 0.6106 | 0.36 | 73 | 0.5439 |
| 0.5766 | 0.72 | 146 | 0.5171 |
| 0.5395 | 1.06 | 219 | 0.5045 |
| 0.5218 | 1.42 | 292 | 0.4976 |
| 0.5336 | 1.78 | 365 | 0.4915 |
| 0.5018 | 2.13 | 438 | 0.4885 |
| 0.5113 | 2.48 | 511 | 0.4856 |
| 0.5066 | 2.84 | 584 | 0.4838 |
| 0.4967 | 3.19 | 657 | 0.4834 |
| 0.4956 | 3.55 | 730 | 0.4830 |
| 0.5026 | 3.9 | 803 | 0.4828 |
Framework versions
- Transformers 4.40.0.dev0
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.0
Quants
Runs of dphn dolphin-2.8-mistral-7b-v02 on huggingface.co
2.9K
Total runs
112
24-hour runs
298
3-day runs
1.2K
7-day runs
2.3K
30-day runs
More Information About dolphin-2.8-mistral-7b-v02 huggingface.co Model
Url of dolphin-2.8-mistral-7b-v02
dolphin-2.8-mistral-7b-v02 huggingface.co Url
Provider of dolphin-2.8-mistral-7b-v02 huggingface.co
Other API from dphn
Updated:September 08 2025
Total runs:
497.7K
Run Growth:
416.1K
Growth Rate:
79.12%
Total runs:
89.2K
Run Growth:
78.8K
Growth Rate:
88.10%
Total runs:
42.7K
Run Growth:
679
Growth Rate:
1.59%
Total runs:
8.9K
Run Growth:
-821
Growth Rate:
-9.24%
Total runs:
8.8K
Run Growth:
-183
Growth Rate:
-2.06%
Total runs:
8.6K
Run Growth:
413
Growth Rate:
4.80%
Total runs:
8.6K
Run Growth:
317
Growth Rate:
3.69%
Total runs:
8.6K
Run Growth:
504
Growth Rate:
5.86%
Total runs:
8.6K
Run Growth:
230
Growth Rate:
2.68%
Total runs:
8.5K
Run Growth:
233
Growth Rate:
2.72%
Total runs:
8.5K
Run Growth:
24
Growth Rate:
0.28%
Total runs:
8.5K
Run Growth:
352
Growth Rate:
4.15%
Total runs:
6.4K
Run Growth:
-694
Growth Rate:
-11.17%
Total runs:
6.3K
Run Growth:
111
Growth Rate:
1.79%
Total runs:
4.5K
Run Growth:
-13.4K
Growth Rate:
-305.59%
Total runs:
3.6K
Run Growth:
-6.6K
Growth Rate:
-189.08%
Total runs:
3.5K
Run Growth:
92
Growth Rate:
2.74%
Total runs:
1.5K
Run Growth:
273
Growth Rate:
18.15%
Total runs:
1.2K
Run Growth:
-1.5K
Growth Rate:
-126.61%
Total runs:
1.2K
Run Growth:
-172
Growth Rate:
-15.47%
Total runs:
1.1K
Run Growth:
-319
Growth Rate:
-27.88%
Total runs:
1.1K
Run Growth:
-11.7K
Growth Rate:
-1012.32%
Total runs:
991
Run Growth:
457
Growth Rate:
46.78%
Total runs:
870
Run Growth:
-68
Growth Rate:
-8.05%
Total runs:
867
Run Growth:
-13
Growth Rate:
-1.52%
Total runs:
808
Run Growth:
-27
Growth Rate:
-3.35%
Total runs:
776
Run Growth:
-1.9K
Growth Rate:
-257.28%
Total runs:
769
Run Growth:
48
Growth Rate:
7.23%
Total runs:
764
Run Growth:
363
Growth Rate:
49.25%
Total runs:
734
Run Growth:
460
Growth Rate:
63.45%
Total runs:
714
Run Growth:
421
Growth Rate:
59.30%
Total runs:
683
Run Growth:
-284
Growth Rate:
-44.58%
Total runs:
592
Run Growth:
104
Growth Rate:
17.63%
Total runs:
503
Run Growth:
-37
Growth Rate:
-6.86%
Total runs:
433
Run Growth:
8
Growth Rate:
1.89%
Total runs:
416
Run Growth:
-149
Growth Rate:
-36.70%
Total runs:
316
Run Growth:
-136
Growth Rate:
-42.37%
Total runs:
225
Run Growth:
-15
Growth Rate:
-6.61%
Total runs:
214
Run Growth:
-196
Growth Rate:
-85.96%
Total runs:
203
Run Growth:
-13
Growth Rate:
-6.47%
Total runs:
184
Run Growth:
119
Growth Rate:
68.39%
Total runs:
175
Run Growth:
-222
Growth Rate:
-124.02%
Total runs:
154
Run Growth:
-227
Growth Rate:
-149.34%
Total runs:
100
Run Growth:
-69
Growth Rate:
-75.82%
Total runs:
100
Run Growth:
-124
Growth Rate:
-125.25%
Total runs:
96
Run Growth:
-34
Growth Rate:
-36.96%
Total runs:
94
Run Growth:
-40
Growth Rate:
-43.01%
Total runs:
74
Run Growth:
37
Growth Rate:
52.86%
Total runs:
71
Run Growth:
39
Growth Rate:
55.71%
Total runs:
70
Run Growth:
24
Growth Rate:
38.71%
Total runs:
70
Run Growth:
48
Growth Rate:
51.61%
Total runs:
69
Run Growth:
12
Growth Rate:
18.18%
Total runs:
62
Run Growth:
39
Growth Rate:
69.64%
Total runs:
50
Run Growth:
18
Growth Rate:
40.91%
Total runs:
38
Run Growth:
-18
Growth Rate:
-54.55%
Total runs:
37
Run Growth:
3
Growth Rate:
10.00%
Total runs:
34
Run Growth:
-19
Growth Rate:
-59.38%
Total runs:
29
Run Growth:
18
Growth Rate:
64.29%
Total runs:
27
Run Growth:
9
Growth Rate:
36.00%
Total runs:
26
Run Growth:
16
Growth Rate:
61.54%
Total runs:
20
Run Growth:
9
Growth Rate:
45.00%
Total runs:
19
Run Growth:
7
Growth Rate:
43.75%
Total runs:
16
Run Growth:
1
Growth Rate:
7.69%
Total runs:
15
Run Growth:
3
Growth Rate:
21.43%
Total runs:
5
Run Growth:
4
Growth Rate:
80.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%
Total runs:
0
Run Growth:
0
Growth Rate:
0.00%