Skip to content
This repository was archived by the owner on Aug 8, 2026. It is now read-only.

Updated metrics.py, exp_ForecastPFN.py, exp_transformer_metalearn.py, requirements.txt - #9

Open
vaidatascientist wants to merge 1 commit into
abacusai:mainfrom
vaidatascientist:vai
Open

Updated metrics.py, exp_ForecastPFN.py, exp_transformer_metalearn.py, requirements.txt#9
vaidatascientist wants to merge 1 commit into
abacusai:mainfrom
vaidatascientist:vai

Conversation

@vaidatascientist

Copy link
Copy Markdown

A) requirements.txt:

  1. pmdarima==2.0.3 --> pmdarima==2.0.4
  2. tensorflow-io==0.26.0 --> tensorflow-io==0.37.1

B) ForecastPFN/benchmark/utils/metrics.py:

  1. from keras import backend --> from tensorflow.keras import backend

C) ForecastPFN/benchmark/exp/exp_ForecastPFN.py:

  1. Casted history variable to float32 to avoid error
  2. Used tf.saved_model.load() instead of tf.keras.models.load_model() as .pb models are not compatible with Keras3
  3. Added code to calculate SMAPE at the bottom

D) ForecastPFN/benchmark/exp/exp_transformer_metalearn.py

  1. Added comments near the imports where changing the sys.path.append() path is important to avoid local development errors

… requirements.txt

A) requirements.txt:
1. pmdarima==2.0.3 --> pmdarima==2.0.4
2. tensorflow-io==0.26.0 --> tensorflow-io==0.37.1

B) ForecastPFN/benchmark/utils/metrics.py:
1. from keras import backend --> from tensorflow.keras import backend

C) ForecastPFN/benchmark/exp/exp_ForecastPFN.py:
1. Casted history variable to float32 to avoid error
2. Used tf.saved_model.load() instead of tf.keras.models.load_model() as .pb models are not compatible with Keras3
3. Added code to calculate SMAPE at the bottom

D) ForecastPFN/benchmark/exp/exp_transformer_metalearn.py
1. Added comments near the imports where changing the sys.path.append() path is important to avoid local development errors
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant