Commit
·
ad85601
1
Parent(s):
b15deaa
Update README.md
Browse files
README.md
CHANGED
|
@@ -11,6 +11,7 @@ tags:
|
|
| 11 |
- llama-2
|
| 12 |
- functions
|
| 13 |
- function calling
|
|
|
|
| 14 |
---
|
| 15 |
# fLlama 2 - Function Calling Llama 2
|
| 16 |
|
|
@@ -18,8 +19,8 @@ tags:
|
|
| 18 |
- The model responds with a structured json argument with the function name and arguments
|
| 19 |
|
| 20 |
Available models:
|
| 21 |
-
- fLlama-7B (
|
| 22 |
-
- [fLlama-13](https://huggingface.co/Trelis/Llama-2-13b-chat-hf-function-calling) - RECOMMENDED. This is a stronger model and handles function calling more accurately.
|
| 23 |
|
| 24 |
## Licensing and Usage
|
| 25 |
|
|
@@ -31,9 +32,9 @@ Some examples of non-commercial use:
|
|
| 31 |
- Personal research
|
| 32 |
|
| 33 |
### Commercial Use
|
| 34 |
-
You can
|
| 35 |
-
|
| 36 |
-
|
| 37 |
|
| 38 |
Some examples of commercial use:
|
| 39 |
- Including or building on fLlama within a product or service, whether monetized or not.
|
|
@@ -41,12 +42,14 @@ Some examples of commercial use:
|
|
| 41 |
|
| 42 |
Commercial license examples:
|
| 43 |
- Running fLlama on a CPU = 1 license required.
|
| 44 |
-
- Running fLlama on a CPU + GPU/TPU = 1 license required. (CPU is considered ancilliary in this case)
|
| 45 |
- Running fLlama sharded across 2 GPUs = 2 licenses required.
|
| 46 |
Licenses are transferable across different machines provided they are not running simultaneously.
|
| 47 |
|
| 48 |
Model weights are provided as-is, without warranty, and with the same disclaimers as provided by Meta (see Llama 2 below).
|
| 49 |
|
|
|
|
|
|
|
| 50 |
### Dataset Licensing
|
| 51 |
|
| 52 |
The dataset used for training this model can be found at [Trelis Function Calling Dataset] (https://huggingface.co/datasets/Trelis/function_calling).
|
|
|
|
| 11 |
- llama-2
|
| 12 |
- functions
|
| 13 |
- function calling
|
| 14 |
+
- sharded
|
| 15 |
---
|
| 16 |
# fLlama 2 - Function Calling Llama 2
|
| 17 |
|
|
|
|
| 19 |
- The model responds with a structured json argument with the function name and arguments
|
| 20 |
|
| 21 |
Available models:
|
| 22 |
+
- [fLlama-7B, 2GB shards](https://huggingface.co/Trelis/Llama-2-7b-chat-hf-function-calling)
|
| 23 |
+
- [fLlama-13, 4GB shards](https://huggingface.co/Trelis/Llama-2-13b-chat-hf-function-calling) - RECOMMENDED. This is a stronger model and handles function calling more accurately.
|
| 24 |
|
| 25 |
## Licensing and Usage
|
| 26 |
|
|
|
|
| 32 |
- Personal research
|
| 33 |
|
| 34 |
### Commercial Use
|
| 35 |
+
You can purchase a life-time license for commercial use:
|
| 36 |
+
- [fLlama-7b: €9.99 per device.](https://buy.stripe.com/7sI14QcWhci71ck3ce)
|
| 37 |
+
- [fLlama-13b: €19.99 per device.](https://buy.stripe.com/9AQ7te3lHdmbdZ68wz)
|
| 38 |
|
| 39 |
Some examples of commercial use:
|
| 40 |
- Including or building on fLlama within a product or service, whether monetized or not.
|
|
|
|
| 42 |
|
| 43 |
Commercial license examples:
|
| 44 |
- Running fLlama on a CPU = 1 license required.
|
| 45 |
+
- Running fLlama on a CPU + 1 GPU/TPU = 1 license required. (CPU is considered ancilliary in this case)
|
| 46 |
- Running fLlama sharded across 2 GPUs = 2 licenses required.
|
| 47 |
Licenses are transferable across different machines provided they are not running simultaneously.
|
| 48 |
|
| 49 |
Model weights are provided as-is, without warranty, and with the same disclaimers as provided by Meta (see Llama 2 below).
|
| 50 |
|
| 51 |
+
*In addition to the above, all use of this model is further subject to terms in the [Meta license](https://ai.meta.com/resources/models-and-libraries/llama-downloads/).
|
| 52 |
+
|
| 53 |
### Dataset Licensing
|
| 54 |
|
| 55 |
The dataset used for training this model can be found at [Trelis Function Calling Dataset] (https://huggingface.co/datasets/Trelis/function_calling).
|