Text Generation
MLX
Safetensors
GGUF
Rust
qwen3_5_text
4b
agentic-coding
android
apple-silicon
attested
bash
c
chain-of-custody
chinese
code
code-completion
code-generation
code-infill
coder
coding
consumer-gpu
cpp
cryptographically-verified
css
delta-forge
edge-inference
embedded
english
forge-alloy
function-calling
ggml
go
html
iphone
java
javascript
kotlin
llama-cpp
lm-studio
local-inference
macbook
mobile
multilingual
ollama
on-device
php
programming
python
q4-k-m
quantized
qwen
qwen3
qwen3.5
raspberry-pi
reproducible
ruby
software-engineering
sql
swift
typescript
Instructions to use continuum-ai/qwen3.5-4b-code-forged with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use continuum-ai/qwen3.5-4b-code-forged with MLX:
# Make sure mlx-lm is installed # pip install --upgrade mlx-lm # if on a CUDA device, also pip install mlx[cuda] # Generate text with mlx-lm from mlx_lm import load, generate model, tokenizer = load("continuum-ai/qwen3.5-4b-code-forged") prompt = "Once upon a time in" text = generate(model, tokenizer, prompt=prompt, verbose=True) - Notebooks
- Google Colab
- Kaggle
- Local Apps
- LM Studio
- MLX LM
How to use continuum-ai/qwen3.5-4b-code-forged with MLX LM:
Generate or start a chat session
# Install MLX LM uv tool install mlx-lm # Generate some text mlx_lm.generate --model "continuum-ai/qwen3.5-4b-code-forged" --prompt "Once upon a time"
Upload forging_results.json with huggingface_hub
Browse files- forging_results.json +14 -14
forging_results.json
CHANGED
|
@@ -2,13 +2,13 @@
|
|
| 2 |
"model": "Qwen/Qwen3.5-4B",
|
| 3 |
"domain": "code",
|
| 4 |
"strategy": "experiential_plasticity",
|
| 5 |
-
"pruning_level": 0.
|
| 6 |
"cycles": 3,
|
| 7 |
-
"training_steps":
|
| 8 |
"baseline_ppl": 3.0382,
|
| 9 |
-
"final_ppl": 2.
|
| 10 |
-
"improvement_pct":
|
| 11 |
-
"forged_at": "2026-03-
|
| 12 |
"device": "NVIDIA GeForce RTX 5090",
|
| 13 |
"tier": "A",
|
| 14 |
"load_4bit": false,
|
|
@@ -20,21 +20,21 @@
|
|
| 20 |
"cycle_results": [
|
| 21 |
{
|
| 22 |
"cycle": 1,
|
| 23 |
-
"post_prune_ppl": 2.
|
| 24 |
-
"post_train_ppl": 2.
|
| 25 |
-
"improvement_vs_baseline_pct":
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"cycle": 2,
|
| 29 |
-
"post_prune_ppl": 2.
|
| 30 |
-
"post_train_ppl": 2.
|
| 31 |
-
"improvement_vs_baseline_pct": 24.
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"cycle": 3,
|
| 35 |
-
"post_prune_ppl": 2.
|
| 36 |
-
"post_train_ppl": 2.
|
| 37 |
-
"improvement_vs_baseline_pct":
|
| 38 |
}
|
| 39 |
],
|
| 40 |
"hardware_targets": [
|
|
|
|
| 2 |
"model": "Qwen/Qwen3.5-4B",
|
| 3 |
"domain": "code",
|
| 4 |
"strategy": "experiential_plasticity",
|
| 5 |
+
"pruning_level": 0.45,
|
| 6 |
"cycles": 3,
|
| 7 |
+
"training_steps": 500,
|
| 8 |
"baseline_ppl": 3.0382,
|
| 9 |
+
"final_ppl": 2.2305,
|
| 10 |
+
"improvement_pct": 26.58,
|
| 11 |
+
"forged_at": "2026-03-28T04:48:47-0500",
|
| 12 |
"device": "NVIDIA GeForce RTX 5090",
|
| 13 |
"tier": "A",
|
| 14 |
"load_4bit": false,
|
|
|
|
| 20 |
"cycle_results": [
|
| 21 |
{
|
| 22 |
"cycle": 1,
|
| 23 |
+
"post_prune_ppl": 2.2001,
|
| 24 |
+
"post_train_ppl": 2.2001,
|
| 25 |
+
"improvement_vs_baseline_pct": 27.59
|
| 26 |
},
|
| 27 |
{
|
| 28 |
"cycle": 2,
|
| 29 |
+
"post_prune_ppl": 2.2839,
|
| 30 |
+
"post_train_ppl": 2.2839,
|
| 31 |
+
"improvement_vs_baseline_pct": 24.83
|
| 32 |
},
|
| 33 |
{
|
| 34 |
"cycle": 3,
|
| 35 |
+
"post_prune_ppl": 2.2305,
|
| 36 |
+
"post_train_ppl": 2.2305,
|
| 37 |
+
"improvement_vs_baseline_pct": 26.58
|
| 38 |
}
|
| 39 |
],
|
| 40 |
"hardware_targets": [
|