Downloads · 30 days
25
5% of all-time downloads
RichardErkhov/wannaphong_-_numfalm-3b-gguf
wannaphong_-_numfalm-3b-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
25
5% of all-time downloads
All-time downloads
525
Public
Repo size
48.5 GB
Likes
0
Public
Click a slice to open those files.
.gguf56.5 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
numfalm-3b - GGUF
| Name | Quant method | Size |
|---|---|---|
| numfalm-3b.Q2_K.gguf | Q2_K | 2.0GB |
| numfalm-3b.IQ3_XS.gguf | IQ3_XS | 2.0GB |
| numfalm-3b.IQ3_S.gguf | IQ3_S | 2.0GB |
| numfalm-3b.Q3_K_S.gguf | Q3_K_S | 2.0GB |
| numfalm-3b.IQ3_M.gguf | IQ3_M | 2.07GB |
| numfalm-3b.Q3_K.gguf | Q3_K | 2.15GB |
| numfalm-3b.Q3_K_M.gguf | Q3_K_M | 2.15GB |
| numfalm-3b.Q3_K_L.gguf | Q3_K_L | 2.22GB |
| numfalm-3b.IQ4_XS.gguf | IQ4_XS | 2.02GB |
| numfalm-3b.Q4_0.gguf | Q4_0 | 2.0GB |
| numfalm-3b.IQ4_NL.gguf | IQ4_NL | 2.02GB |
| numfalm-3b.Q4_K_S.gguf | Q4_K_S | 2.41GB |
| numfalm-3b.Q4_K.gguf | Q4_K | 2.57GB |
| numfalm-3b.Q4_K_M.gguf | Q4_K_M | 2.57GB |
| numfalm-3b.Q4_1.gguf | Q4_1 | 2.2GB |
| numfalm-3b.Q5_0.gguf | Q5_0 | 2.4GB |
| numfalm-3b.Q5_K_S.gguf | Q5_K_S | 2.6GB |
| numfalm-3b.Q5_K.gguf | Q5_K | 2.74GB |
| numfalm-3b.Q5_K_M.gguf | Q5_K_M | 2.74GB |
| numfalm-3b.Q5_1.gguf | Q5_1 | 2.6GB |
| numfalm-3b.Q6_K.gguf | Q6_K | 3.6GB |
| numfalm-3b.Q8_0.gguf | Q8_0 | 3.6GB |
license: apache-2.0 language:
NumFaLM 3B is a bilingual language model trained in Thai and English. The architecture model is Llama model that pretraining from scratch. It was built to open source AI and research for bilingual language models and improve small language models. We released the training script and train datasets so you can research the training and datasets.
We fork EasyLM and added training by HuggingFace datasets, but HuggingFace was down many times during the time we trained the model, so we can train just one epoch. The model trained one epoch.
Research supported with Cloud TPUs from Google's TPU Research Cloud (TRC). We use TPU4-64 for training model about 4 days / 1 epoch.
Thank you TPU Research Cloud and EasyLM project! We use EasyLM for pretraining model.