Model card: general description, IRx-2 Best Overall, use-case examples

README.md

@@ -10,24 +10,46 @@ tags:
 - on-device
 ---
 
-# IRx-2 Gemma
+# IRx-2 Pro
 
-The IRx-2 fine-tune on a Gemma 4 E4B base: a private, offline chat assistant for
-Apple Silicon (this MLX repo) and phones/tablets
-([GGUF build](https://huggingface.co/ikppramesh/irx-2-gemma-GGUF)).
+The largest edition of the IRx family: a private, offline AI chat assistant for long,
+detailed answers and extended conversations. Everything runs on your device — no
+internet connection, no account, and nothing you type ever leaves it.
 
-| Model | MLX (Apple Silicon) | GGUF (PocketPal, LM Studio, llama.cpp) |
-|---|---|---|
-| **IRx-2 Gemma** | [ikppramesh/irx-2-gemma](https://huggingface.co/ikppramesh/irx-2-gemma) | [ikppramesh/irx-2-gemma-GGUF](https://huggingface.co/ikppramesh/irx-2-gemma-GGUF) |
-| IRx-2 (4B) | [ikppramesh/irx-2](https://huggingface.co/ikppramesh/irx-2) | [ikppramesh/irx-2-GGUF](https://huggingface.co/ikppramesh/irx-2-GGUF) |
-| IRx-1 (2B, fastest) | [ikppramesh/irx-1](https://huggingface.co/ikppramesh/irx-1) | [ikppramesh/irx-1-GGUF](https://huggingface.co/ikppramesh/irx-1-GGUF) |
+## The IRx family
 
-Same training data and pipeline as IRx-1/IRx-2, trained with a 2048-token window so
-it learned from nearly all long conversations in the dataset. Thinking mode is always
-off (answers start immediately), identity is built in, and builds are only published
-after an automated check for looping and repeated answers in multi-turn chats.
+All three run fully offline — nothing you type ever leaves your device.
 
-## Usage
+| Model | Download | Best for | Phones & tablets (GGUF) | Mac (MLX) |
+|---|---|---|---|---|
+| **IRx-1** | ~1.2GB | The fastest. Quick everyday answers on almost any phone | [irx-1-GGUF](https://huggingface.co/ikppramesh/irx-1-GGUF) | [irx-1](https://huggingface.co/ikppramesh/irx-1) |
+| **⭐ IRx-2 — Best Overall** | ~2.7GB | The best balance of answer quality and speed | [irx-2-GGUF](https://huggingface.co/ikppramesh/irx-2-GGUF) | [irx-2](https://huggingface.co/ikppramesh/irx-2) |
+| **IRx-2 Pro** | ~5.3GB | The most detailed, long-form answers and long conversations | [irx-2-pro-GGUF](https://huggingface.co/ikppramesh/irx-2-pro-GGUF) | [irx-2-pro](https://huggingface.co/ikppramesh/irx-2-pro) |
+
+**Not sure which to pick? Start with ⭐ IRx-2.** Choose IRx-1 for speed or an older /
+smaller phone, and IRx-2 Pro for the richest answers on a device with 12GB+ RAM.
+
+## What IRx-2 Pro does best
+
+IRx-2 Pro is the largest edition. It was trained with the longest context window of
+the family (so it learned from nearly all of the long conversations in the training
+data), which makes it the best choice for long, detailed answers and for working
+through a topic over many messages. It runs at about the same speed as IRx-2, but
+needs more memory: use it on devices with 12GB+ RAM, or on a Mac.
+
+| Who | Example things to ask |
+|---|---|
+| Software developers | "Help me design a small inventory app step by step — data model, screens and API." · "Walk me through refactoring this code over a few messages." |
+| Farmers | "Help me write a full-season plan for my farm — sowing, watering, spraying and harvest — then let's refine it." · "Write a detailed proposal to sell my produce to a local supermarket." |
+| Students & researchers | "Turn these notes into a structured report with headings and a summary." · "Explain this topic in depth, then quiz me on it." |
+| Teachers | "Create a full week's unit on fractions: daily lessons, activities and a final test." |
+| Business owners | "Draft a detailed marketing plan for my new bakery." · "Write a complete employee handbook outline." |
+| Writers & creators | "Write a 1,500-word blog post on healthy habits." · "Help me develop a short story over several messages." |
+| Everyday life | "Plan a 7-day family trip with a day-by-day schedule and budget." |
+
+Measured on the same Mac CPU: IRx-2 Pro generates about 24 tokens/s, close to IRx-2's 26 — most of its extra size is lookup tables that cost little compute.
+
+## Usage (Mac, MLX)
 
 ```bash
 pip install mlx-lm
@@ -37,25 +59,41 @@ pip install mlx-lm
 from mlx_lm import load, generate
 from mlx_lm.sample_utils import make_sampler
 
-model, tokenizer = load("ikppramesh/irx-2-gemma")
+model, tokenizer = load("ikppramesh/irx-2-pro")
 messages = [{"role": "user", "content": "How do I convert Celsius to Fahrenheit?"}]
 prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
 print(generate(model, tokenizer, prompt=prompt, max_tokens=1024,
                sampler=make_sampler(temp=0.7, top_p=0.95)))
 ```
 
-No system prompt is required; the built-in template supplies the IRx-2 one when none
-is given.
+No system prompt is required: the built-in chat template supplies the IRx-2 Pro one when
+none is given, and always keeps thinking mode off. Sample at a non-zero temperature
+(e.g. `temp=0.7`). **On a phone or tablet**, use the
+[GGUF build](https://huggingface.co/ikppramesh/irx-2-pro-GGUF).
 
 ## Limitations
 
-- **Not a frontier-scale model** — won't match large hosted models on hard multi-step
-  reasoning, deep coding problems, or breadth of knowledge.
-- **Facts can be wrong** — double-check specific facts, as with any small model.
-- **Not current-events aware** — knowledge is fixed at training time.
-- **Don't enable native tool/function-calling in chat apps** — never trained.
+- **Not a frontier-scale model** — it won't match large hosted AI services on very hard
+  multi-step reasoning, deep coding problems or breadth of world knowledge.
+- **Facts can be wrong.** Like any small offline model it can state things confidently
+  that aren't true — double-check anything important, such as prices, medicines and
+  doses, laws, and local farming advice (seed varieties, chemical quantities, weather).
+- **Not a replacement for professionals** — medical, legal and financial answers are
+  general information only.
+- **Not current-events aware** — its knowledge is fixed at training time.
+- **Don't enable native tool/function-calling in chat apps** — never trained; plain
+  chat is reliable.
+
+## Changelog
+
+- **2026-09-30** — First release (published earlier the same day under a different
+  repository name; old links redirect here). Trained with a 2048-token window, merged
+  into the full-precision base and quantized once with an importance matrix. Passed the
+  automated pre-publish check: repeated 8-turn conversations with the repeat penalty off
+  and the app's thinking request on — no loops, no copied answers, no hidden reasoning,
+  correct identity.
 
 ## License
 
-Apache 2.0. IRx-2 Gemma is a derivative fine-tuned model — full Apache 2.0 terms
-apply as with any work under this license.
+Apache 2.0. IRx-2 Pro is a derivative fine-tuned model — full Apache 2.0 terms apply as with
+any work under this license.