File size: 7,975 Bytes

---
base_model: Writer/Palmyra-Fin-70B-32K
extra_gated_fields:
  Email: text
  I acknowledge that this model is for non-commercial use only unless I acquire a separate license from Writer: checkbox
  Name: text
  Organization or Affiliation: text
  Receive email updates and promotions on Writer products, services, and research?:
    options:
    - 'Yes'
    - 'No'
    type: select
extra_gated_prompt: By clicking "Agree", you agree to the [License Agreement](https://writer.com/legal/open-model-license/)
  and acknowledge Writer's [Privacy Policy](https://writer.com/legal/acceptable-use/).
inference: false
language:
- en
library_name: gguf
license: other
license_link: https://writer.com/legal/open-model-license/
license_name: writer-open-model-license
model-index:
- name: Palmyra-Fin-70B-32k
  results: []
pipeline_tag: text-generation
quantized_by: legraphista
tags:
- instruct
- pytorch
- finance
- stock market
- candlesticks
- FinGPT
- option trading
- future stock prediction
- trends prediction
- Enterprise LLM
- Enterprise
- Enterprise ready
- Banks
- Wealth Management
- quantized
- GGUF
- quantization
- imat
- imatrix
- static
- 32bit
- 16bit
- 8bit
- 6bit
- 5bit
- 4bit
- 3bit
- 2bit
- 1bit
---

# Palmyra-Fin-70B-32K-IMat-GGUF
_Llama.cpp imatrix quantization of Writer/Palmyra-Fin-70B-32K_

Original Model: [Writer/Palmyra-Fin-70B-32K](https://huggingface.co/Writer/Palmyra-Fin-70B-32K)    
Original dtype: `BF16` (`bfloat16`)  
Quantized by:  llama.cpp [b3504](https://github.com/ggerganov/llama.cpp/releases/tag/b3504)  
IMatrix dataset: [here](https://gist.githubusercontent.com/bartowski1182/eb213dccb3571f863da82e99418f81e8/raw/b2869d80f5c16fd7082594248e80144677736635/calibration_datav3.txt)  

- [Files](#files)
    - [IMatrix](#imatrix)
    - [Common Quants](#common-quants)
    - [All Quants](#all-quants)
- [Downloading using huggingface-cli](#downloading-using-huggingface-cli)
- [Inference](#inference)
    - [Simple chat template](#simple-chat-template)
    - [Chat template with system prompt](#chat-template-with-system-prompt)
    - [Llama.cpp](#llama-cpp)
- [FAQ](#faq)
    - [Why is the IMatrix not applied everywhere?](#why-is-the-imatrix-not-applied-everywhere)
    - [How do I merge a split GGUF?](#how-do-i-merge-a-split-gguf)

---

## Files

### IMatrix
Status: ✅ Available  
Link: [here](https://huggingface.co/legraphista/Palmyra-Fin-70B-32K-IMat-GGUF/blob/main/imatrix.dat)

### Common Quants
| Filename | Quant type | File Size | Status | Uses IMatrix | Is Split |
| -------- | ---------- | --------- | ------ | ------------ | -------- |
| [Palmyra-Fin-70B-32K.Q8_0/*](https://huggingface.co/legraphista/Palmyra-Fin-70B-32K-IMat-GGUF/tree/main/Palmyra-Fin-70B-32K.Q8_0) | Q8_0 | 74.98GB | ✅ Available | ⚪ Static | ✂ Yes
| [Palmyra-Fin-70B-32K.Q6_K/*](https://huggingface.co/legraphista/Palmyra-Fin-70B-32K-IMat-GGUF/tree/main/Palmyra-Fin-70B-32K.Q6_K) | Q6_K | 57.89GB | ✅ Available | ⚪ Static | ✂ Yes
| Palmyra-Fin-70B-32K.Q4_K | Q4_K | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q3_K | Q3_K | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q2_K | Q2_K | - | ⏳ Processing | 🟢 IMatrix | -


### All Quants
| Filename | Quant type | File Size | Status | Uses IMatrix | Is Split |
| -------- | ---------- | --------- | ------ | ------------ | -------- |
| [Palmyra-Fin-70B-32K.F32/*](https://huggingface.co/legraphista/Palmyra-Fin-70B-32K-IMat-GGUF/tree/main/Palmyra-Fin-70B-32K.F32) | F32 | 282.22GB | ✅ Available | ⚪ Static | ✂ Yes
| Palmyra-Fin-70B-32K.BF16 | BF16 | - | ⏳ Processing | ⚪ Static | -
| Palmyra-Fin-70B-32K.FP16 | F16 | - | ⏳ Processing | ⚪ Static | -
| [Palmyra-Fin-70B-32K.Q8_0/*](https://huggingface.co/legraphista/Palmyra-Fin-70B-32K-IMat-GGUF/tree/main/Palmyra-Fin-70B-32K.Q8_0) | Q8_0 | 74.98GB | ✅ Available | ⚪ Static | ✂ Yes
| [Palmyra-Fin-70B-32K.Q6_K/*](https://huggingface.co/legraphista/Palmyra-Fin-70B-32K-IMat-GGUF/tree/main/Palmyra-Fin-70B-32K.Q6_K) | Q6_K | 57.89GB | ✅ Available | ⚪ Static | ✂ Yes
| Palmyra-Fin-70B-32K.Q5_K | Q5_K | - | ⏳ Processing | ⚪ Static | -
| Palmyra-Fin-70B-32K.Q5_K_S | Q5_K_S | - | ⏳ Processing | ⚪ Static | -
| Palmyra-Fin-70B-32K.Q4_K | Q4_K | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q4_K_S | Q4_K_S | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ4_NL | IQ4_NL | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ4_XS | IQ4_XS | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q3_K | Q3_K | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q3_K_L | Q3_K_L | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q3_K_S | Q3_K_S | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ3_M | IQ3_M | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ3_S | IQ3_S | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ3_XS | IQ3_XS | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ3_XXS | IQ3_XXS | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q2_K | Q2_K | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.Q2_K_S | Q2_K_S | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ2_M | IQ2_M | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ2_S | IQ2_S | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ2_XS | IQ2_XS | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ2_XXS | IQ2_XXS | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ1_M | IQ1_M | - | ⏳ Processing | 🟢 IMatrix | -
| Palmyra-Fin-70B-32K.IQ1_S | IQ1_S | - | ⏳ Processing | 🟢 IMatrix | -


## Downloading using huggingface-cli
If you do not have hugginface-cli installed:
```
pip install -U "huggingface_hub[cli]"
```
Download the specific file you want:
```
huggingface-cli download legraphista/Palmyra-Fin-70B-32K-IMat-GGUF --include "Palmyra-Fin-70B-32K.Q8_0.gguf" --local-dir ./
```
If the model file is big, it has been split into multiple files. In order to download them all to a local folder, run:
```
huggingface-cli download legraphista/Palmyra-Fin-70B-32K-IMat-GGUF --include "Palmyra-Fin-70B-32K.Q8_0/*" --local-dir ./
# see FAQ for merging GGUF's
```

---

## Inference

### Simple chat template
```
<|begin_of_text|><|start_header_id|>user<|end_header_id|>

{user_prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

{assistant_response}<|eot_id|><|start_header_id|>user<|end_header_id|>

{next_user_prompt}<|eot_id|>
```

### Chat template with system prompt
```
<|begin_of_text|><|start_header_id|>system<|end_header_id|>

{system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>

{user_prompt}<|eot_id|><|start_header_id|>assistant<|end_header_id|>

{assistant_response}<|eot_id|><|start_header_id|>user<|end_header_id|>

{next_user_prompt}<|eot_id|>
```

### Llama.cpp
```
llama.cpp/main -m Palmyra-Fin-70B-32K.Q8_0.gguf --color -i -p "prompt here (according to the chat template)"
```

---

## FAQ

### Why is the IMatrix not applied everywhere?
According to [this investigation](https://www.reddit.com/r/LocalLLaMA/comments/1993iro/ggufs_quants_can_punch_above_their_weights_now/), it appears that lower quantizations are the only ones that benefit from the imatrix input (as per hellaswag results). 

### How do I merge a split GGUF?
1. Make sure you have `gguf-split` available
    - To get hold of `gguf-split`, navigate to https://github.com/ggerganov/llama.cpp/releases
    - Download the appropriate zip for your system from the latest release
    - Unzip the archive and you should be able to find `gguf-split`
2. Locate your GGUF chunks folder (ex: `Palmyra-Fin-70B-32K.Q8_0`)
3. Run `gguf-split --merge Palmyra-Fin-70B-32K.Q8_0/Palmyra-Fin-70B-32K.Q8_0-00001-of-XXXXX.gguf Palmyra-Fin-70B-32K.Q8_0.gguf`
    - Make sure to point `gguf-split` to the first chunk of the split.

---

Got a suggestion? Ping me [@legraphista](https://x.com/legraphista)!