prithivMLmods/crm-01-4b-f32-GGUF

crm-01-4b-f32-GGUF

Arc-Intelligence/crm-01-4b is a specialized, Qwen3-4B-based language model fine-tuned for high-stakes CRM (Customer Relationship Management) policy tasks such as quote approvals and invalid configuration identification. Its training focused on tool-aware instruction following, enabling it to call helper functions (as defined in crm_sandbox/env/functions.py) and return concise, policy-compliant answers. Enhanced with domain-specific skills and a learning orchestrator methodology, it demonstrates substantial performance improvements on policy-driven CRM challenges—most notably, a 32.1% gain in overall policy task success and a 57% improvement on quote approval benchmarks. Designed primarily for enterprise automation, crm-01-4b excels at handling procedural workflows and approvals in simulated Salesforce environments, but is less suited for general open-ended conversation or domains beyond CRM without further fine-tuning. Users can integrate it within agent frameworks or via Hugging Face Transformers for streamlined policy execution in business settings.

Model Files

Quants Usage

(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)

Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):