You can deploy this model and start earning money today!

RichardErkhov/
HuggingFaceTB_-_SmolLM-1.7B-Instruct-gguf

Name: HuggingFaceTB_-_SmolLM-1.7B-Instruct-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

SmolLM-1.7B-Instruct - GGUF

Model creator: https://huggingface.co/HuggingFaceTB/
Original model: https://huggingface.co/HuggingFaceTB/SmolLM-1.7B-Instruct/

Name	Quant method	Size
SmolLM-1.7B-Instruct.Q2_K.gguf	Q2_K	0.63GB
SmolLM-1.7B-Instruct.IQ3_XS.gguf	IQ3_XS	0.69GB
SmolLM-1.7B-Instruct.IQ3_S.gguf	IQ3_S	0.72GB
SmolLM-1.7B-Instruct.Q3_K_S.gguf	Q3_K_S	0.72GB
SmolLM-1.7B-Instruct.IQ3_M.gguf	IQ3_M	0.75GB
SmolLM-1.7B-Instruct.Q3_K.gguf	Q3_K	0.8GB
SmolLM-1.7B-Instruct.Q3_K_M.gguf	Q3_K_M	0.8GB
SmolLM-1.7B-Instruct.Q3_K_L.gguf	Q3_K_L	0.87GB
SmolLM-1.7B-Instruct.IQ4_XS.gguf	IQ4_XS	0.88GB
SmolLM-1.7B-Instruct.Q4_0.gguf	Q4_0	0.92GB
SmolLM-1.7B-Instruct.IQ4_NL.gguf	IQ4_NL	0.93GB
SmolLM-1.7B-Instruct.Q4_K_S.gguf	Q4_K_S	0.93GB
SmolLM-1.7B-Instruct.Q4_K.gguf	Q4_K	0.98GB
SmolLM-1.7B-Instruct.Q4_K_M.gguf	Q4_K_M	0.98GB
SmolLM-1.7B-Instruct.Q4_1.gguf	Q4_1	1.02GB
SmolLM-1.7B-Instruct.Q5_0.gguf	Q5_0	1.11GB
SmolLM-1.7B-Instruct.Q5_K_S.gguf	Q5_K_S	1.11GB
SmolLM-1.7B-Instruct.Q5_K.gguf	Q5_K	1.14GB
SmolLM-1.7B-Instruct.Q5_K_M.gguf	Q5_K_M	1.14GB
SmolLM-1.7B-Instruct.Q5_1.gguf	Q5_1	1.2GB
SmolLM-1.7B-Instruct.Q6_K.gguf	Q6_K	1.31GB
SmolLM-1.7B-Instruct.Q8_0.gguf	Q8_0	1.7GB

Original model description:

license: apache-2.0 base_model: HuggingFaceTB/SmolLM-1.7B tags:

alignment-handbook
trl
sft datasets:
Magpie-Align/Magpie-Pro-300K-Filtered
bigcode/self-oss-instruct-sc2-exec-filter-50k
teknium/OpenHermes-2.5
HuggingFaceTB/everyday-conversations-llama3.1-2k library_name: transformers language:
en

SmolLM-1.7B-Instruct

Model Summary

SmolLM is a series of small language models available in three sizes: 135M, 360M, and 1.7B parameters.

These models are pre-trained on SmolLM-Corpus, a curated collection of high-quality educational and synthetic data designed for training LLMs. For further details, we refer to our blogpost.

To build SmolLM-Instruct, we finetuned the base models on publicly available datasets.

Changelog

Release	Description
v0.1	Initial release of SmolLM-Instruct. We finetune on the permissive subset of the WebInstructSub dataset, combined with StarCoder2-Self-OSS-Instruct. Then, we perform DPO (Direct Preference Optimization) for one epoch on HelpSteer for the 135M and 1.7B models, and argilla/dpo-mix-7k for the 360M model.
v0.2	We changed the finetuning mix to datasets more suitable for smol models. We train on a new dataset of 2k simple everyday conversations we generated by llama3.1-70B everyday-conversations-llama3.1-2k, Magpie-Pro-300K-Filtered, StarCoder2-Self-OSS-Instruct, and a small subset of OpenHermes-2.5

v0.2 models are better at staying on topic and responding appropriately to standard prompts, such as greetings and questions about their role as AI assistants. SmolLM-360M-Instruct (v0.2) has a 63.3% win rate over SmolLM-360M-Instruct (v0.1) on AlpacaEval. You can find the details here.

You can load v0.1 checkpoint by specifying revision="v0.1" in the transformers code:

model = AutoModelForCausalLM.from_pretrained("HuggingFaceTB/SmolLM-1.7B-Instruct", revision="v0.1")

Usage

Local Applications

⚡ For local applications, you can find optimized implementations of the model in MLC, GGUF and Transformers.js formats, in addition to fast in-browser demos in this collection: https://huggingface.co/collections/HuggingFaceTB/local-smollms-66c0f3b2a15b4eed7fb198d0

We noticed that 4bit quantization degrades the quality of the 135M and 360M, so we use q016 for MLC and ONNX/Transformers.js checkpoints for the WebGPU demos. We also suggest using temperature 0.2 and top-p 0.9.

Transformers

pip install transformers

python

# pip install transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
checkpoint = "HuggingFaceTB/SmolLM-1.7B-Instruct"

device = "cuda" # for GPU usage or "cpu" for CPU usage
tokenizer = AutoTokenizer.from_pretrained(checkpoint)
# for multiple GPUs install accelerate and do `model = AutoModelForCausalLM.from_pretrained(checkpoint, device_map="auto")`
model = AutoModelForCausalLM.from_pretrained(checkpoint).to(device)

messages = [{"role": "user", "content": "What is the capital of France."}]
input_text=tokenizer.apply_chat_template(messages, tokenize=False)
print(input_text)
inputs = tokenizer.encode(input_text, return_tensors="pt").to(device)
outputs = model.generate(inputs, max_new_tokens=50, temperature=0.2, top_p=0.9, do_sample=True)
print(tokenizer.decode(outputs[0]))

Chat in TRL

You can also use the TRL CLI to chat with the model from the terminal:

bash

pip install trl
trl chat --model_name_or_path HuggingFaceTB/SmolLM-1.7B-Instruct --device cpu

Limitations

Additionally, the generated content may not always be factually accurate, logically consistent, or free from biases present in the training data, we invite users to leverage them as assistive tools rather than definitive sources of information. We find that they can handle general knowledge questions, creative writing and basic Python programming. But they are English only and may have difficulty with arithmetics, editing tasks and complex reasoning. For more details about the models' capabilities, please refer to our blog post.

Training parameters

We train the models using the alignment-handbook with the datasets mentioned in the changelog, using these parameters v0.2 (most of them are from Zephyr Gemma recipe):

1 epoch
lr 1e-3
cosine schedule
warmup ratio 0.1
global batch size 262k tokens

You can find the training recipe here: https://github.com/huggingface/alignment-handbook/tree/smollm/recipes/smollm

Citation