Downloads · 30 days
23
4% of all-time downloads
RichardErkhov/bhenrym14_-_platypus-yi-34b-gguf
bhenrym14_-_platypus-yi-34b-gguf is a machine learning model from RichardErkhov. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
Downloads · 30 days
23
4% of all-time downloads
All-time downloads
557
Public
Repo size
377 GB
Likes
0
Public
Click a slice to open those files.
.gguf414 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
platypus-yi-34b - GGUF
| Name | Quant method | Size |
|---|---|---|
| platypus-yi-34b.Q2_K.gguf | Q2_K | 11.94GB |
| platypus-yi-34b.IQ3_XS.gguf | IQ3_XS | 2.26GB |
| platypus-yi-34b.IQ3_S.gguf | IQ3_S | 13.99GB |
| platypus-yi-34b.Q3_K_S.gguf | Q3_K_S | 13.93GB |
| platypus-yi-34b.IQ3_M.gguf | IQ3_M | 4.11GB |
| platypus-yi-34b.Q3_K.gguf | Q3_K | 15.51GB |
| platypus-yi-34b.Q3_K_M.gguf | Q3_K_M | 15.51GB |
| platypus-yi-34b.Q3_K_L.gguf | Q3_K_L | 16.89GB |
| platypus-yi-34b.IQ4_XS.gguf | IQ4_XS | 17.36GB |
| platypus-yi-34b.Q4_0.gguf | Q4_0 | 18.13GB |
| platypus-yi-34b.IQ4_NL.gguf | IQ4_NL | 18.3GB |
| platypus-yi-34b.Q4_K_S.gguf | Q4_K_S | 18.25GB |
| platypus-yi-34b.Q4_K.gguf | Q4_K | 19.24GB |
| platypus-yi-34b.Q4_K_M.gguf | Q4_K_M | 19.24GB |
| platypus-yi-34b.Q4_1.gguf | Q4_1 | 20.1GB |
| platypus-yi-34b.Q5_0.gguf | Q5_0 | 22.08GB |
| platypus-yi-34b.Q5_K_S.gguf | Q5_K_S | 22.08GB |
| platypus-yi-34b.Q5_K.gguf | Q5_K | 22.65GB |
| platypus-yi-34b.Q5_K_M.gguf | Q5_K_M | 15.12GB |
| platypus-yi-34b.Q5_1.gguf | Q5_1 | 24.05GB |
| platypus-yi-34b.Q6_K.gguf | Q6_K | 23.37GB |
| platypus-yi-34b.Q8_0.gguf | Q8_0 | 31.88GB |
license: other license_name: yi-license license_link: LICENSE datasets:
This is chargoddard/Yi-34B-Llama, with instruction tuning performed with the garage-bAInd/Open-Platypus dataset. That base model is 01-ai/Yi-34B, but using llama2 model definitions and tokenizer to remove any remote code requirements.
This is a (merged) QLoRA fine-tune (rank 64).
The finetune was performed with 1x RTX 6000 Ada (~18 hours to this checkpoint). It is possible this is rather undertrained, as this checkpoint is at 1 epoch. I began to see some performance degradation after that; more hyperparameter tuning is probably warranted.
Use as you would any llama-2 model.
Model was trained with legacy airoboros <2.0 system prompt. See bhenrym14/airoboros-33b-gpt4-1.4.1-lxctx-PI-16384-fp16 model card for details.