RonanMcGovern commited on
Commit
08e6d8e
·
1 Parent(s): b1ea2c6

update videos and products

Browse files

Trelis Research provides tools and tutorials for training, using and deploying large language models.

# Most-viewed Video Tutorials [updated weekly]

1. [Run Llama 2 on your Laptop with Jupyter](https://www.youtube.com/watch?v=nDJMHFsBU7M)
2. [Code Llama Tutorial](https://www.youtube.com/watch?v=rjSWCMVbD_U)
3. [A Detailed Guide to Fine tuning Language Models with QLoRa](https://www.youtube.com/watch?v=OQdp-OeG1as)

# Top Selling Products [updated weekly]

1. [Function-calling Llama 2/Code-Llama](https://huggingface.co/Trelis/Llama-2-13b-chat-hf-function-calling-v2)
2. [Advanced QLoRA Training Script / Fine-tuning Colab Notebook](https://buy.stripe.com/5kA5l69K52Hxf3a006)
3. [Long Document and Website Summarisation Tool](https://Summarise-Me.com)

Files changed (1) hide show
  1. README.md +10 -30
README.md CHANGED
@@ -5,39 +5,19 @@ colorFrom: red
5
  colorTo: green
6
  sdk: static
7
  pinned: false
8
- tags: [llama, jupyter, colab, function calling, QLoRa, fine-tuning, scripts, fine-tuning]
9
  ---
10
 
11
- # 1. QLoRa Fine-Tuning Scripts/Notebooks
12
 
13
- **Basic Training Notebook**
14
- - Access the Google Colab script [here](https://colab.research.google.com/drive/1uMSS1o_8YOPyG1X_4k6ENEE3kJfBGGhH?usp=sharing).
15
 
16
- - View the [Youtube Video Here](https://youtu.be/OQdp-OeG1as)
 
 
17
 
18
- **ADVANCED Training Notebook (PAID)**
19
- This advanced script provides improved performance when training with small datasets:
20
- - Includes a prompt loss-mask for improved performance when structured responses are required.
21
- - Includes a stop token after responses - allowing the model to provide a short reponse (e.g. a function call) and then stop.
22
- - Request [access here](https://buy.stripe.com/5kA5l69K52Hxf3a006). €14.99 (or $16.49) per seat/user. Access will be provided within 24 hours of purchase.
23
 
24
- # 2. Datasets
25
-
26
- *Function Calling Dataset for Fine-Tuning*
27
- - Commercial dataset allowing language models to be fine-tuned for function calling ([Paid access here](https://huggingface.co/datasets/Trelis/function_calling_extended)).
28
- - Created using only human input or Apache 2 licensed datasets (no third party commercial licensing limitations)
29
- - Models trained with this dataset:
30
- - - [Llama-2-7B](https://huggingface.co/Trelis/Llama-2-7b-chat-hf-function-calling-v2) - repo includes Google Colab notebook for inference
31
- - - [Llama-2-13B](https://huggingface.co/Trelis/Llama-2-13b-chat-hf-function-calling-v2) - repo includes Google Colab notebook for inference
32
-
33
- *Protein Stability*
34
- - 250k dataset of protein mutations and their effect on stability. [Free access here](https://huggingface.co/datasets/Trelis/protein_stability_single_mutation).
35
-
36
- # 3. Jupyter Llama for Laptop
37
-
38
- A Chat Assistant built on Llama 2.
39
- - Save and re-load chats.
40
- - Upload pdf or text files for analysis.
41
- - Search Bing (beta).
42
- - Add custom functions (beta).
43
- - Get access for €9.99 ($10.99) [here](https://buy.stripe.com/dR65l6f4p95V7AI6oA).
 
5
  colorTo: green
6
  sdk: static
7
  pinned: false
8
+ tags: [llama, code-llama, jupyter, colab, function calling, QLoRa, fine-tuning, scripts, fine-tuning]
9
  ---
10
 
11
+ Trelis Research provides tools and tutorials for training, using and deploying large language models.
12
 
13
+ # Most-viewed Video Tutorials [updated weekly]
 
14
 
15
+ 1. [Run Llama 2 on your Laptop with Jupyter](https://www.youtube.com/watch?v=nDJMHFsBU7M)
16
+ 2. [Code Llama Tutorial](https://www.youtube.com/watch?v=rjSWCMVbD_U)
17
+ 3. [A Detailed Guide to Fine tuning Language Models with QLoRa](https://www.youtube.com/watch?v=OQdp-OeG1as)
18
 
19
+ # Top Selling Products [updated weekly]
 
 
 
 
20
 
21
+ 1. [Function-calling Llama 2/Code-Llama Model](https://huggingface.co/Trelis/Llama-2-13b-chat-hf-function-calling-v2)
22
+ 2. [Advanced QLoRA Training Script / Fine-tuning Colab Notebook - see video #3 above](https://buy.stripe.com/5kA5l69K52Hxf3a006)
23
+ 3. [Long Document and Website Summarisation Tool](https://Summarise-Me.com)