RichardErkhov/rogercc_-_LLama3-8B-CraftaxClassic-LoRa-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

LLama3-8B-CraftaxClassic-LoRa - GGUF

• Model creator: https://huggingface.co/rogercc/
• Original model: https://huggingface.co/rogercc/LLama3-8B-CraftaxClassic-LoRa/

Name	Quant method	Size
LLama3-8B-CraftaxClassic-LoRa.Q2_K.gguf	Q2_K	2.96GB
LLama3-8B-CraftaxClassic-LoRa.IQ3_XS.gguf	IQ3_XS	3.28GB
LLama3-8B-CraftaxClassic-LoRa.IQ3_S.gguf	IQ3_S	3.43GB
LLama3-8B-CraftaxClassic-LoRa.Q3_K_S.gguf	Q3_K_S	3.41GB
LLama3-8B-CraftaxClassic-LoRa.IQ3_M.gguf	IQ3_M	3.52GB
LLama3-8B-CraftaxClassic-LoRa.Q3_K.gguf	Q3_K	3.74GB
LLama3-8B-CraftaxClassic-LoRa.Q3_K_M.gguf	Q3_K_M	3.74GB
LLama3-8B-CraftaxClassic-LoRa.Q3_K_L.gguf	Q3_K_L	4.03GB
LLama3-8B-CraftaxClassic-LoRa.IQ4_XS.gguf	IQ4_XS	4.18GB
LLama3-8B-CraftaxClassic-LoRa.Q4_0.gguf	Q4_0	4.34GB
LLama3-8B-CraftaxClassic-LoRa.IQ4_NL.gguf	IQ4_NL	4.38GB
LLama3-8B-CraftaxClassic-LoRa.Q4_K_S.gguf	Q4_K_S	4.37GB
LLama3-8B-CraftaxClassic-LoRa.Q4_K.gguf	Q4_K	4.58GB
LLama3-8B-CraftaxClassic-LoRa.Q4_K_M.gguf	Q4_K_M	4.58GB
LLama3-8B-CraftaxClassic-LoRa.Q4_1.gguf	Q4_1	4.78GB
LLama3-8B-CraftaxClassic-LoRa.Q5_0.gguf	Q5_0	5.21GB
LLama3-8B-CraftaxClassic-LoRa.Q5_K_S.gguf	Q5_K_S	5.21GB
LLama3-8B-CraftaxClassic-LoRa.Q5_K.gguf	Q5_K	5.34GB
LLama3-8B-CraftaxClassic-LoRa.Q5_K_M.gguf	Q5_K_M	5.34GB
LLama3-8B-CraftaxClassic-LoRa.Q5_1.gguf	Q5_1	5.65GB
LLama3-8B-CraftaxClassic-LoRa.Q6_K.gguf	Q6_K	6.14GB
LLama3-8B-CraftaxClassic-LoRa.Q8_0.gguf	Q8_0	7.95GB

Original model description: LLama3-8B fine-tuned on data collected by the PPO baseline (after training) on Craftax-Classic. This model can be used to obtain promising actions at a given state.

---

license: mit