Downloads · 30 days
6
8% of all-time downloads
SmolLM2-1.7B-Scheduled-QAT-Gradual-INT4
jpcurada/SmolLM2-1.7B-Scheduled-QAT-Gradual-INT4
SmolLM2-1.7B-Scheduled-QAT-Gradual-INT4 is a machine learning model from jpcurada. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
This model was produced via Hardware-Aware Scheduled Quantization-Aware Training using the Gradual snapping strategy, trained on WikiText-103 with a Kaggle TPU v5e-8.
All-time downloads
75
Public
Parameters
1.7B
3.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors3.4 GB · 100%
At a glance
- License
- apache-2.0
- Model type
- llama
- Access
- Public
- Created
- May 9, 2026
- Updated
- May 9, 2026
- SHA
- b6cdfca8
Base models
Datasets
Model card
From the Hugging Face model README
SmolLM2-1.7B -- Scheduled QAT (Gradual) INT4
This model was produced via Hardware-Aware Scheduled Quantization-Aware Training using the Gradual snapping strategy, trained on WikiText-103 with a Kaggle TPU v5e-8.
Results
| Metric | Value |
|---|---|
| Test Perplexity | 26.7619 |
| Test Loss | 3.2870 |
| KL Divergence | 0.949187 nats |
| Training Steps | 3,821 |
| Training Time | 14,581s (4.1 hours) |
Strategy Description
Gradual [24, 16, 8]: Balanced transition strategy with equal-duration phases across all bit-widths.
- Continuous 24-32 -> INT32
- Continuous 16-24 -> INT16
- Continuous 8-16 -> INT8
- Continuous 4-8 -> INT4
This creates a smooth, balanced progression allowing the model to adapt gradually to each quantization level.
Usage
HuggingFace Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained(
"jpcurada/SmolLM2-1.7B-Scheduled-QAT-Gradual-INT4"
)
tokenizer = AutoTokenizer.from_pretrained(
"jpcurada/SmolLM2-1.7B-Scheduled-QAT-Gradual-INT4"
)
GGUF (llama.cpp)
llama-cli -m smollm2-1.7b-sched-qat-gradual-Q4_K_M.gguf -p "Hello, world!"
Training Details
- Base model: SmolLM2-1.7B
- Dataset: WikiText-103 (244,597 train sequences)
- Sequence length: 512
- Effective batch size: 64 (1 x 8 grad-accum x 8 TPU cores)
- Optimizer: Adafactor (lr=2e-5)
- LR Schedule: Cosine annealing
- Hardware: TPU v5e-8 (bfloat16)
- Type
- llama
- License
- apache-2.0
- Created
- May 9, 2026
- Updated
- May 9, 2026