README.md

---
library_name: peft
base_model: meta-llama/Llama-3.1-8B-Instruct
pipeline_tag: text-generation
language:
- en
license: cc-by-nc-4.0
tags:
- lora
- peft
- molly-os
- specialist
- climate-analytics-manager
- llama-3.1
- domain-adaptation
---

# Molly Specialist — Climate Analytics Manager

Quantifies Scope 1-3 emissions across portfolios, benchmarks against SBTi pathways, and translates climate scenarios into financial risk metrics more accurately than the base model.

Part of **[Molly](https://iamolly.ai/?utm_source=huggingface&utm_medium=model_card&utm_campaign=specialists&utm_content=molly-climate-analytics-manager)**, an orchestrator that keeps a library of small domain
specialists over one quantized base and routes each request to the right one, so a
single machine answers across many fields without loading a separate large model
for each.

## What this specialist handles well

- Calculates financed emissions using PCAF methodology across asset classes
- Maps physical climate hazards to asset-level financial exposure with TCFD alignment
- Compares transition scenarios from NGFS and IEA for portfolio stress testing

## Try it with

- "What are the PCAF emission factors for corporate equity holdings in 2026?"
- "How do NGFS Net Zero 2050 scenario assumptions affect our oil and gas portfolio?"
- "Which assets in our real estate portfolio face highest acute physical risk by 2030?"

## Before you run: the base model is gated

This adapter needs the base weights, and the base is **access-gated**. Do this **once**:

1. Accept the base licence: <https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct>
2. Create a **read token**: <https://huggingface.co/settings/tokens>
3. Make the token available:
   - **Google Colab:** Secrets panel (key icon) → *Add new secret* → name `HF_TOKEN`, enable **Notebook access**.
   - **Kaggle:** *Add-ons → Secrets* → add `HF_TOKEN`.
   - **Local:** `huggingface-cli login` or `export HF_TOKEN=...`

Skipping this gives `GatedRepoError` / `401 Unauthorized` when the **base** loads. A stored
Colab secret is **not** applied automatically — authenticate in code, as below.

## Quickstart

```python
# pip install -U transformers peft accelerate
import os, torch
from huggingface_hub import login
try:
    from google.colab import userdata
    login(userdata.get("HF_TOKEN"))
except Exception:
    tok = os.environ.get("HF_TOKEN")
    login(tok) if tok else login()

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

BASE = "meta-llama/Llama-3.1-8B-Instruct"
ADAPTER = "BoomJules/molly-climate-analytics-manager"

tok = AutoTokenizer.from_pretrained(BASE)
base = AutoModelForCausalLM.from_pretrained(BASE, torch_dtype=torch.bfloat16, device_map="auto")
model = PeftModel.from_pretrained(base, ADAPTER).eval()

msgs = [{"role": "user", "content": "Your question here"}]
ids = tok.apply_chat_template(msgs, add_generation_prompt=True, return_tensors="pt").to(model.device)
out = model.generate(ids, max_new_tokens=300)
print(tok.decode(out[0][ids.shape[1]:], skip_special_tokens=True))
```

## Low-VRAM (4-bit) — fits a free Colab/Kaggle GPU (~6–7 GB)

```python
# pip install -U transformers peft accelerate bitsandbytes
import os, torch
from huggingface_hub import login
try:
    from google.colab import userdata
    login(userdata.get("HF_TOKEN"))
except Exception:
    login(os.environ.get("HF_TOKEN"))

from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
from peft import PeftModel

bnb = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_quant_type="nf4",
                         bnb_4bit_compute_dtype=torch.bfloat16, bnb_4bit_use_double_quant=True)
tok = AutoTokenizer.from_pretrained("meta-llama/Llama-3.1-8B-Instruct")
base = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.1-8B-Instruct", quantization_config=bnb, device_map="auto")
model = PeftModel.from_pretrained(base, "BoomJules/molly-climate-analytics-manager").eval()
```

## Adapter details

| | |
|---|---|
| Base model | `meta-llama/Llama-3.1-8B-Instruct` |
| Method | LoRA (PEFT) |
| Rank / alpha | 32 / 64 |
| Domain | Climate Analytics Manager |

## Troubleshooting

- **`GatedRepoError` / `401 Unauthorized`** — base licence not accepted, or `HF_TOKEN` missing,
  or the Colab secret was stored but `login(...)` was never called.
- **CUDA out of memory** — use the 4-bit snippet on a GPU runtime.
- **Adapter seems to have no effect** — confirm the base id matches `base_model` above.

## Other Molly specialists

- [Quantum Software Architect](https://huggingface.co/BoomJules/molly-quantum-software-architect)
- [Quantum Communication Systems Engineer](https://huggingface.co/BoomJules/molly-quantum-communication-systems-engineer)
- [Infectious Disease Physician Antimicrobial Stewardship](https://huggingface.co/BoomJules/molly-infectious-disease-physician-antimicrobial-stewardship)
- [Health Informatics Medical AI Specialist](https://huggingface.co/BoomJules/molly-health-informatics-medical-ai-specialist)
- [Clinical Trial Pharmacologist](https://huggingface.co/BoomJules/molly-clinical-trial-pharmacologist)
- [Immunopharmacologist](https://huggingface.co/BoomJules/molly-immunopharmacologist)
- [Language Technology Consultant](https://huggingface.co/BoomJules/molly-language-technology-consultant)
- [Polymer Chemist](https://huggingface.co/BoomJules/molly-polymer-chemist)
- [Composite Materials Engineer](https://huggingface.co/BoomJules/molly-composite-materials-engineer)
- [Computer Science AI](https://huggingface.co/BoomJules/molly-cs-ai)
- [Computer Science Algorithms](https://huggingface.co/BoomJules/molly-cs-algorithms)
- [Computer Science Computer Vision](https://huggingface.co/BoomJules/molly-cs-cv)

Running several of these at once, with the routing decided for you, is what
[Molly](https://iamolly.ai/?utm_source=huggingface&utm_medium=model_card&utm_campaign=specialists&utm_content=molly-climate-analytics-manager) does.

## Licence & intended use

Adapter: **CC BY-NC 4.0** (attribution, non-commercial). Base model: its own licence.
Intended for research and evaluation in Climate Analytics Manager.

© 2026 Core Labs R&D.