Downloads · 30 days
7
9% of all-time downloads
pocketai-jarvis-intent-1.7b
Umangs28/pocketai-jarvis-intent-1.7b
pocketai-jarvis-intent-1.7b is a machine learning model from Umangs28. 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.
Base: unsloth/Qwen2.5-1.5B-Instruct (4-bit prepacked) — 1.5B params Training: SFT + LoRA via Unsloth, trainonresponsesonly masking Data: ~3000 utterance→JSON intent pairs (Hinglish + English mix) Quantization: Q4KM, ~…
All-time downloads
82
Public
Repo size
986 MB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.gguf986 MB · 100%
At a glance
- License
- apache-2.0
- Access
- Public
- Created
- May 17, 2026
- Updated
- May 17, 2026
- SHA
- 0d5776f6
Base models
Model card
From the Hugging Face model README
PocketAI JARVIS Intent Parser (1.5B Q4_K_M)
Base: unsloth/Qwen2.5-1.5B-Instruct (4-bit prepacked) — 1.5B params
Training: SFT + LoRA via Unsloth, train_on_responses_only masking
Data: ~3000 utterance→JSON intent pairs (Hinglish + English mix)
Quantization: Q4_K_M, ~934 MiB on disk, 5.08 BPW
Format: GGUF, llama.cpp compatible
Schema
{"action": "OPEN_APP|CALL|MESSAGE|SET_ALARM|SET_TIMER|SEARCH_WEB|SEARCH_IN_APP|PLAY_MEDIA|TAKE_PHOTO|TOGGLE_FLASHLIGHT|SET_VOLUME|SET_BRIGHTNESS|SCROLL|GO_BACK|CLEAR_TEXT|TAP_ELEMENT|TYPE_TEXT|UNKNOWN",
"app": "string?",
"target": "string?",
"text": "string?",
"time": "string?",
"query": "string?"}
Use in PocketAI
Drop-in for Tier 2 of IntentPipeline (replaces the previously disabled LlmIntentParser).
System prompt expects model to output ONLY the JSON object — no markdown, no prose.
- License
- apache-2.0
- Created
- May 17, 2026
- Updated
- May 17, 2026