Spark-X2.6-4B-AgentFast is a fine-tuned checkpoint of XHToken/Spark-X2.6-4B tuned specifically for autonomous tool calling, system operations, and prompt compliance without deliberation latency.
Empirical "Happy Accident": Our original fine-tuning objective was to train the model on full multi-step Fable deliberation traces. However, due to a chat-template field mapping mismatch where
<thought>traces were not parsed into the template's expectedreasoning_contentkey, the training targets were prepared with empty reasoning channels (<|Bot|></think>).Over 150 steps of SFT, the model inadvertently learned an aggressive "early-exit" reflex: it slams the thinking channel shut at token position 0 and immediately emits tool calls and direct answers. Because the underlying transformer weights remained intact, it preserved its core instruction and math skills while running 30–40% faster. We are releasing this artifact as an interesting low-latency edge agent case study while actively preparing a corrected fine-tune that fully preserves reasoning traces.
Evaluated against the stock post-trained Spark-X2.6-4B instruct model across three official benchmark suites on Vulkan1 (Radeon PRO V320):
exec_simple)Evaluated on 30 official test samples with strict Python AST syntax and argument verification:
Evaluated with Google Research's canonical test suite:
Evaluated on 15 objective multiple-choice questions from the 2026 Gaokao Math exam to verify reasoning retention:
reasoning_effort=low. AgentFast immediately closes the thought channel (</think>) and emits direct tool calls (<tool_call>) and structured responses in 2–4 seconds (~61 tokens/sec on Vulkan).<arg_key>) and JSON argument payloads ready for live executor pipelines without hesitations or conversational preambles.Quantized GGUF files are available at JamieBradfield/Spark-X2.6-4B-AgentFast-GGUF.
CLI One-Liner (Vulkan / ROCm / CUDA / CPU):
llama-cli \
-m Spark-X2.6-Fable-Step150-IQ3_M.gguf \
-p "<|start▁of▁sentence|><|User|>List 3 core rules of memory management.<|end▁of▁sentence|><|start▁of▁sentence|><|Bot|>" \
-n 512 \
-ngl 99 \
-c 4096 \
--temp 0.2 \
--single-turn
Serve as OpenAI-Compatible API:
llama-server \
-m Spark-X2.6-Fable-Step150-IQ3_M.gguf \
--port 8080 \
-ngl 99 \
-c 4096
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
model_id = "JamieBradfield/Spark-X2.6-4B-AgentFast"
tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
model_id,
trust_remote_code=True,
torch_dtype=torch.bfloat16,
device_map="auto"
)
messages = [{"role": "user", "content": "List 3 core rules of memory management."}]
inputs = tokenizer.apply_chat_template(messages, tokenize=True, add_generation_prompt=True, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=512, temperature=0.2)
print(tokenizer.decode(outputs[0][inputs.shape[-1]:], skip_special_tokens=True))
model.safetensors with embedded Spark2_5 architecture files.Spark-X2.6-Fable-Step150-IQ3_M.gguf (1.94 GB, fits in ~2.1 GB VRAM at 1024 context).Notice: This model is shared as an experimental research release provided "as-is", without warranties of any kind, express or implied. There is no commitment or warranty of future updates, maintenance, or downstream bug fixes.
Have the files? Submit a torrent.
Submit a torrentHave the complete files and permission to share them? Replace the example folder path. The script checks the files, makes a torrent and submits its magnet and checksums. Keep your torrent client seeding. Get your community key from your settings, and read the script before running it.
curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo JamieBradfield/Spark-X2.6-4B-AgentFast ./model-folderXHToken/Spark-X2.6-4B