Add pipeline tag and link to project page
Browse filesThis PR adds the text-generation pipeline tag, enabling people to find this model at https://huggingface.co/models?pipeline_tag=text-generation as well as adds the missing project page.
README.md
CHANGED
|
@@ -1,14 +1,16 @@
|
|
| 1 |
---
|
| 2 |
-
library_name: transformers
|
| 3 |
language:
|
| 4 |
- en
|
|
|
|
| 5 |
license: cc-by-nc-4.0
|
|
|
|
| 6 |
---
|
| 7 |
|
| 8 |
# Model Information
|
| 9 |
|
| 10 |
We introduce **UltraLong-8B**, a series of ultra-long context language models designed to process extensive sequences of text (up to 1M, 2M, and 4M tokens) while maintaining competitive performance on standard benchmarks. Built on the Llama-3.1, UltraLong-8B leverages a systematic training recipe that combines efficient continued pretraining with instruction tuning to enhance long-context understanding and instruction-following capabilities. This approach enables our models to efficiently scale their context windows without sacrificing general performance.
|
| 11 |
|
|
|
|
| 12 |
|
| 13 |
## The UltraLong Models
|
| 14 |
|
|
|
|
| 1 |
---
|
|
|
|
| 2 |
language:
|
| 3 |
- en
|
| 4 |
+
library_name: transformers
|
| 5 |
license: cc-by-nc-4.0
|
| 6 |
+
pipeline_tag: text-generation
|
| 7 |
---
|
| 8 |
|
| 9 |
# Model Information
|
| 10 |
|
| 11 |
We introduce **UltraLong-8B**, a series of ultra-long context language models designed to process extensive sequences of text (up to 1M, 2M, and 4M tokens) while maintaining competitive performance on standard benchmarks. Built on the Llama-3.1, UltraLong-8B leverages a systematic training recipe that combines efficient continued pretraining with instruction tuning to enhance long-context understanding and instruction-following capabilities. This approach enables our models to efficiently scale their context windows without sacrificing general performance.
|
| 12 |
|
| 13 |
+
Project page: https://ultralong.github.io/
|
| 14 |
|
| 15 |
## The UltraLong Models
|
| 16 |
|