rishdotblog (Rishabh) – Community Activity

New activity in defog/llama-3-sqlcoder-8b 7 months ago

#19 opened 7 months ago by

#2 opened 9 months ago by

New activity in defog/sqlcoder-70b-alpha 11 months ago

#12 opened 11 months ago by

New activity in defog/sqlcoder-7b-2 11 months ago

#18 opened 11 months ago by

#13 opened 11 months ago by

#14 opened 11 months ago by

#10 opened 12 months ago by

#12 opened 11 months ago by

New activity in defog/sqlcoder-7b-2 12 months ago

#9 opened 12 months ago by

#8 opened 12 months ago by

#7 opened 12 months ago by

#2 opened 12 months ago by

#3 opened 12 months ago by

#4 opened 12 months ago by

#5 opened 12 months ago by

#6 opened 12 months ago by

New activity in defog/sqlcoder-70b-alpha about 1 year ago

#6 opened about 1 year ago by

#2 opened about 1 year ago by

#5 opened about 1 year ago by

#4 opened about 1 year ago by

Rishabh

rishdotblog's activity

Rishabh

AI & ML interests

Organizations

rishdotblog's activity

Remove sampling and set temperature to zero.

Replace the `eos_token` with `<eot_id>`, to prevent model from going on unnecessarily long

Expected inputs Azure ML Studio

SQL Query always points to PostgreSQL, even if strongly prompted not to do

context length higher than 100K

Fine-tuning with SQL coder

为什么生成的like查询语句总是包含ILIKE

How did this 7b model get such high accuracy

how to use sqlcoder-7b-q5_k_m.gguf in llama.cpp

Supplementary data

Amazon Sagemaker endpoint implementation doesn't work accurately.

Prompt in the model card is slightly different from the one provided in Github

What SQL dialects are supported?

Is using the database schema the best way to describe prime the model?

Dates

Context-Window

Not able to run the with 4bit quantization on RTX4090 (24GB vRAM) local GPU

Does not work

Excuse me, is the current model a chat model?

Can't host with vLLM - "LlamaModel" architecture is not supported.