Downloads · 30 days
28
1% of all-time downloads
RichardErkhov/Kendamarron_-_Tokara-0.5B-v0.1-gguf
Kendamarron_-_Tokara-0.5B-v0.1-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
28
1% of all-time downloads
All-time downloads
2.6K
Public
Repo size
6.1 GB
Likes
0
Public
Click a slice to open those files.
.gguf7 GB · 100%
From the Hugging Face model README
Quantization made by Richard Erkhov.
Tokara-0.5B-v0.1 - GGUF
| Name | Quant method | Size |
|---|---|---|
| Tokara-0.5B-v0.1.Q2_K.gguf | Q2_K | 0.23GB |
| Tokara-0.5B-v0.1.IQ3_XS.gguf | IQ3_XS | 0.24GB |
| Tokara-0.5B-v0.1.IQ3_S.gguf | IQ3_S | 0.25GB |
| Tokara-0.5B-v0.1.Q3_K_S.gguf | Q3_K_S | 0.25GB |
| Tokara-0.5B-v0.1.IQ3_M.gguf | IQ3_M | 0.26GB |
| Tokara-0.5B-v0.1.Q3_K.gguf | Q3_K | 0.26GB |
| Tokara-0.5B-v0.1.Q3_K_M.gguf | Q3_K_M | 0.26GB |
| Tokara-0.5B-v0.1.Q3_K_L.gguf | Q3_K_L | 0.28GB |
| Tokara-0.5B-v0.1.IQ4_XS.gguf | IQ4_XS | 0.28GB |
| Tokara-0.5B-v0.1.Q4_0.gguf | Q4_0 | 0.29GB |
| Tokara-0.5B-v0.1.IQ4_NL.gguf | IQ4_NL | 0.29GB |
| Tokara-0.5B-v0.1.Q4_K_S.gguf | Q4_K_S | 0.29GB |
| Tokara-0.5B-v0.1.Q4_K.gguf | Q4_K | 0.3GB |
| Tokara-0.5B-v0.1.Q4_K_M.gguf | Q4_K_M | 0.3GB |
| Tokara-0.5B-v0.1.Q4_1.gguf | Q4_1 | 0.3GB |
| Tokara-0.5B-v0.1.Q5_0.gguf | Q5_0 | 0.32GB |
| Tokara-0.5B-v0.1.Q5_K_S.gguf | Q5_K_S | 0.32GB |
| Tokara-0.5B-v0.1.Q5_K.gguf | Q5_K | 0.33GB |
| Tokara-0.5B-v0.1.Q5_K_M.gguf | Q5_K_M | 0.33GB |
| Tokara-0.5B-v0.1.Q5_1.gguf | Q5_1 | 0.34GB |
| Tokara-0.5B-v0.1.Q6_K.gguf | Q6_K | 0.36GB |
| Tokara-0.5B-v0.1.Q8_0.gguf | Q8_0 | 0.47GB |
license: other license_name: tongyi-qianwen-research license_link: https://huggingface.co/Qwen/Qwen1.5-0.5B/blob/main/LICENSE language:
Qwen/Qwen1.5-0.5Bを日英データ5Bトークンで継続事前学習したモデルです。
ベンチマークのスコアは低下していますが、ベースモデルよりも安定して日本語を出力するようになっています。
詳細はこちらをご覧ください。
Stability-AI/lm-evaluation-harnessの3項目で評価
| モデル | jsquad(1-shot) | jcommonsenseqa(1-shot) | jnli(1-shot) |
|---|---|---|---|
| Kendamarron/Tokara-0.5B-v0.1 | 26.4295 | 0.2663 | 0.5509 |
| Qwen/Qwen1.5-0.5B | 31.3597 | 0.2556 | 0.5534 |
日本の在来馬であるトカラ馬から
from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
model = AutoModelForCausalLM.from_pretrained('Kendamarron/Tokara-0.5B-v0.1')
tokenizer = AutoTokenizer.from_pretrained('Kendamarron/Tokara-0.5B-v0.1')
pipe = pipeline('text-generation', model=model, tokenizer=tokenizer)
prompt = "大規模言語モデルとは、"
print(pipe(prompt, max_length=128, repetition_penalty=1.1, temperature=0.7, top_p=0.95))