RichardErkhov/SpectraSuite_-_FloatLM_2.4B-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

FloatLM_2.4B - GGUF

• Model creator: https://huggingface.co/SpectraSuite/
• Original model: https://huggingface.co/SpectraSuite/FloatLM_2.4B/

Name	Quant method	Size
FloatLM_2.4B.Q2_K.gguf	Q2_K	0.89GB
FloatLM_2.4B.IQ3_XS.gguf	IQ3_XS	0.98GB
FloatLM_2.4B.IQ3_S.gguf	IQ3_S	1.03GB
FloatLM_2.4B.Q3_K_S.gguf	Q3_K_S	1.03GB
FloatLM_2.4B.IQ3_M.gguf	IQ3_M	1.08GB
FloatLM_2.4B.Q3_K.gguf	Q3_K	1.14GB
FloatLM_2.4B.Q3_K_M.gguf	Q3_K_M	1.14GB
FloatLM_2.4B.Q3_K_L.gguf	Q3_K_L	1.23GB
FloatLM_2.4B.IQ4_XS.gguf	IQ4_XS	1.26GB
FloatLM_2.4B.Q4_0.gguf	Q4_0	1.32GB
FloatLM_2.4B.IQ4_NL.gguf	IQ4_NL	1.33GB
FloatLM_2.4B.Q4_K_S.gguf	Q4_K_S	1.33GB
FloatLM_2.4B.Q4_K.gguf	Q4_K	1.4GB
FloatLM_2.4B.Q4_K_M.gguf	Q4_K_M	1.4GB
FloatLM_2.4B.Q4_1.gguf	Q4_1	1.46GB
FloatLM_2.4B.Q5_0.gguf	Q5_0	1.59GB
FloatLM_2.4B.Q5_K_S.gguf	Q5_K_S	1.59GB
FloatLM_2.4B.Q5_K.gguf	Q5_K	1.63GB
FloatLM_2.4B.Q5_K_M.gguf	Q5_K_M	1.63GB
FloatLM_2.4B.Q5_1.gguf	Q5_1	1.73GB
FloatLM_2.4B.Q6_K.gguf	Q6_K	1.88GB
FloatLM_2.4B.Q8_0.gguf	Q8_0	2.44GB

Original model description:

license: apache-2.0

FloatLM 2.4B

The good ol' FP16 LLMs with LLaMa architecture.

import transformers as tf, torch
model_name = "SpectraSuite/FloatLM_2.4B"

# Please adjust the temperature, repetition penalty, top_k, top_p and other sampling parameters according to your needs.
pipeline = tf.pipeline("text-generation", model=model_id, model_kwargs={"torch_dtype": torch.float16}, device_map="auto")

# These are base (pretrained) LLMs that are not instruction and chat tuned. You may need to adjust your prompt accordingly.
pipeline("Once upon a time")

• License: Apache 2.0
• We will use our GitHub repo for communication (including HF repo related queries). Feel free to open an issue here https://github.com/NolanoOrg/SpectraSuite