somosnlp-hackathon-2025/Auxiliar_investigacion_papers
Updated
AI & ML interests
None defined yet.
Recent Activity
View all activity
jferrinm
updated
a
model
about 1 month ago
jferrinm
published
a
model
about 1 month ago
mariagrandury
authored
4
papers
2 months ago
Adding LLMs to the psycholinguistic norming toolbox: A practical guide to getting the most out of human ratings
Paper
•
2509.14405
•
Published
•
2
Psycholinguistic Word Features: a New Approach for the Evaluation of LLMs Alignment with Humans
Paper
•
2506.22439
•
Published
•
3
Apertus: Democratizing Open and Compliant LLMs for Global Language Environments
Paper
•
2509.14233
•
Published
•
12
La Leaderboard: A Large Language Model Leaderboard for Spanish Varieties and Languages of Spain and Latin America
Paper
•
2507.00999
•
Published
•
1
dianags
authored
2
papers
6 months ago
mariagrandury
authored
2
papers
6 months ago
pcuenq
authored
a
paper
8 months ago
angelika
authored
3
papers
12 months ago
Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation
Paper
•
2412.03304
•
Published
•
21
MEDITRON-70B: Scaling Medical Pretraining for Large Language Models
Paper
•
2311.16079
•
Published
•
19
CRAB: Assessing the Strength of Causal Relationships Between Real-world Events
Paper
•
2311.04284
•
Published
ouhenio
authored
2
papers
about 1 year ago
mariagrandury
authored
a
paper
about 1 year ago
mariagrandury
authored
2
papers
over 1 year ago
Post
9875
OpenELM in Core ML
Apple recently released a set of efficient LLMs in sizes varying between 270M and 3B parameters. Their quality, according to benchmarks, is similar to OLMo models of comparable size, but they required half the pre-training tokens because they use layer-wise scaling, where the number of attention heads increases in deeper layers.
I converted these models to Core ML, for use on Apple Silicon, using this script: https://gist.github.com/pcuenca/23cd08443460bc90854e2a6f0f575084. The converted models were uploaded to this community in the Hub for anyone that wants to integrate inside their apps: corenet-community/openelm-core-ml-6630c6b19268a5d878cfd194
The conversion was done with the following parameters:
- Precision: float32.
- Sequence length: fixed to 128.
With swift-transformers (https://github.com/huggingface/swift-transformers), I'm getting about 56 tok/s with the 270M on my M1 Max, and 6.5 with the largest 3B model. These speeds could be improved by converting to
I'm also looking at optimizing inference using an experimental kv cache in swift-transformers. It's a bit tricky because the layers have varying number of attention heads, but I'm curious to see how much this feature can accelerate performance in this model family :)
Regarding the instruct fine-tuned models, I don't know the chat template that was used. The models use the Llama 2 tokenizer, but the Llama 2 chat template, or the default Alignment Handbook one that was used to train, are not recognized. Any ideas on this welcome!
Apple recently released a set of efficient LLMs in sizes varying between 270M and 3B parameters. Their quality, according to benchmarks, is similar to OLMo models of comparable size, but they required half the pre-training tokens because they use layer-wise scaling, where the number of attention heads increases in deeper layers.
I converted these models to Core ML, for use on Apple Silicon, using this script: https://gist.github.com/pcuenca/23cd08443460bc90854e2a6f0f575084. The converted models were uploaded to this community in the Hub for anyone that wants to integrate inside their apps: corenet-community/openelm-core-ml-6630c6b19268a5d878cfd194
The conversion was done with the following parameters:
- Precision: float32.
- Sequence length: fixed to 128.
With swift-transformers (https://github.com/huggingface/swift-transformers), I'm getting about 56 tok/s with the 270M on my M1 Max, and 6.5 with the largest 3B model. These speeds could be improved by converting to
float16. However, there's some precision loss somewhere and generation doesn't work in float16 mode yet. I'm looking into this and will keep you posted! Or take a look at this issue if you'd like to help: https://github.com/huggingface/swift-transformers/issues/95I'm also looking at optimizing inference using an experimental kv cache in swift-transformers. It's a bit tricky because the layers have varying number of attention heads, but I'm curious to see how much this feature can accelerate performance in this model family :)
Regarding the instruct fine-tuned models, I don't know the chat template that was used. The models use the Llama 2 tokenizer, but the Llama 2 chat template, or the default Alignment Handbook one that was used to train, are not recognized. Any ideas on this welcome!
-
- 5 replies