Downloads · 30 days
51
4% of all-time downloads
RichardErkhov/M4-ai_-_neural-chat-mini-v2.2-1.8B-gguf
M4-ai_-_neural-chat-mini-v2.2-1.8B-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
51
4% of all-time downloads
All-time downloads
1.4K
Public
Repo size
19.8 GB
Likes
0
Public
Click a slice to open those files.
.gguf23.4 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
neural-chat-mini-v2.2-1.8B - GGUF
| Name | Quant method | Size |
|---|---|---|
| neural-chat-mini-v2.2-1.8B.Q2_K.gguf | Q2_K | 0.79GB |
| neural-chat-mini-v2.2-1.8B.Q3_K_S.gguf | Q3_K_S | 0.89GB |
| neural-chat-mini-v2.2-1.8B.Q3_K.gguf | Q3_K | 0.95GB |
| neural-chat-mini-v2.2-1.8B.Q3_K_M.gguf | Q3_K_M | 0.95GB |
| neural-chat-mini-v2.2-1.8B.Q3_K_L.gguf | Q3_K_L | 0.98GB |
| neural-chat-mini-v2.2-1.8B.IQ4_XS.gguf | IQ4_XS | 1.01GB |
| neural-chat-mini-v2.2-1.8B.Q4_0.gguf | Q4_0 | 1.04GB |
| neural-chat-mini-v2.2-1.8B.IQ4_NL.gguf | IQ4_NL | 1.05GB |
| neural-chat-mini-v2.2-1.8B.Q4_K_S.gguf | Q4_K_S | 1.08GB |
| neural-chat-mini-v2.2-1.8B.Q4_K.gguf | Q4_K | 1.13GB |
| neural-chat-mini-v2.2-1.8B.Q4_K_M.gguf | Q4_K_M | 1.13GB |
| neural-chat-mini-v2.2-1.8B.Q4_1.gguf | Q4_1 | 1.13GB |
| neural-chat-mini-v2.2-1.8B.Q5_0.gguf | Q5_0 | 1.22GB |
| neural-chat-mini-v2.2-1.8B.Q5_K_S.gguf | Q5_K_S | 1.24GB |
| neural-chat-mini-v2.2-1.8B.Q5_K.gguf | Q5_K | 1.28GB |
| neural-chat-mini-v2.2-1.8B.Q5_K_M.gguf | Q5_K_M | 1.28GB |
| neural-chat-mini-v2.2-1.8B.Q5_1.gguf | Q5_1 | 1.31GB |
| neural-chat-mini-v2.2-1.8B.Q6_K.gguf | Q6_K | 1.47GB |
| neural-chat-mini-v2.2-1.8B.Q8_0.gguf | Q8_0 | 1.82GB |
language:
We fine-tuned tau-1.8B using SFT and DPOP on a high quality mix for general-purpose assistants.
This model has capabilities in math, coding, writing, and more. We fine-tuned it using a high quality mix for general-purpose assistants.
General purpose assistant, question answering, chain-of-thought, etc..
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
coming soon
We used 8 Kaggle TPUs, and we trained at a global batch size of 128 and sequence length of 2048.