初始化项目,由ModelHub XC社区提供模型

Model: uralstech/kcc-krishi-qwen3-1.7b
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-10-13 19:13:16 +08:00
commit 9a28e3a2f7
12 changed files with 739 additions and 0 deletions

121
LICENSE.md Normal file
View File

@@ -0,0 +1,121 @@
# KCC-Krishi Qwen3-1.7B — Licensing and Attribution Notes
## License metadata
The Hugging Face model repository uses:
```yaml
license: other
```
This is intentional because the release is a derived model combining:
- the Qwen/Qwen3-1.7B base model;
- KCC-derived training evidence;
- KCC-Krishi fine-tuning;
- project-specific model packaging;
- project-specific runtime safety components;
- source and derived-data attribution requirements.
## Base model
This model is derived from:
```text
Qwen/Qwen3-1.7B
```
Users must review and comply with the base model license and attribution terms before using, modifying, or redistributing this derived model.
The base model license remains separate from the KCC-Krishi dataset and project-processing terms.
## Dataset attribution
The training evidence is derived from:
**Kisan Call Centre (KCC) Transcripts of Farmers Queries and Answers**
Open Government Data Platform India / Data.gov.in
Ministry of Agriculture and Farmers Welfare, Government of India
Accessed through AIKosh/Data.gov.in
The KCC-Krishi processing, translation, quality control, routing, fine-tuning, packaging, and runtime safeguards were created by:
```text
KCC-Krishi Project Team
URAV ADVANCED LEARNING SYSTEMS PRIVATE LIMITED
```
## Derived model terms
This model is provided for research, reproducibility, RAG, SFT, evaluation, and controlled offline-assistant prototyping.
Users must:
- preserve Qwen base-model attribution;
- preserve KCC source attribution;
- preserve KCC-Krishi attribution;
- retain this licensing note;
- comply with applicable source-data and model-license requirements;
- review current agricultural regulations and official guidance;
- use runtime safety and evidence-grounding components for deployment.
## Prohibited interpretation
This model must not be presented as:
- field-validated agronomic ground truth;
- an unrestricted agricultural authority;
- a replacement for agronomists;
- a current weather or market information source;
- a government-scheme status authority;
- an autonomous pesticide or fertilizer recommendation system;
- a substitute for product-label or local-agriculture-officer verification.
## Runtime requirement
The intended deployment includes:
```text
out-of-scope guard
→ current-information guard
→ regional RAG retrieval
→ Qwen model
→ agrochemical exactness guard
→ final safety validation
```
Deployment without these controls is outside the intended project design.
## No endorsement
No endorsement is implied by:
- the Government of India;
- Data.gov.in;
- AIKosh;
- the Ministry of Agriculture and Farmers Welfare;
- Kisan Call Centre;
- Qwen;
- Hugging Face;
- Microsoft Azure;
- URAV Advanced Learning Systems Private Limited.
## No warranty
The model is provided without a warranty that:
- every generated answer is correct;
- every source-derived recommendation is current;
- every product, formulation, unit, or dose is correct;
- every response is suitable for a particular crop or location;
- the model is safe for unattended field deployment;
- the model will not reproduce source-derived text.
## Safety notice
Pesticide, fertilizer, disease, weather, market, and government-scheme information must be independently checked against current official sources, current product labels, local regulations, and qualified agricultural advice.
## Legal notice
This document is not legal advice. Users are responsible for evaluating licensing, attribution, privacy, redistribution, safety, and regulatory obligations before using or distributing the model.