Dominik Moritz Roth dodox
dodox pushed to master at dodox/itpal_jax 2025-01-07 18:24:49 +01:00
1096dbd848 Perf tests
404320c5cc revert kl, cxant kit compile c-binding
Compare 2 commits »
dodox pushed to master at dodox/itpal_jax 2025-01-07 16:54:40 +01:00
4d6ed9b3ac Better jit (bool mask via matmul)
dodox pushed to master at dodox/itpal_jax 2024-12-21 19:21:40 +01:00
7fca6186d5 jit wherever possible
dodox pushed to master at dodox/itpal_jax 2024-12-21 18:53:52 +01:00
2e0ca977bc Update README
e83cb9a8a5 Also check loss calc works for full cov case
3e2b988a2f Fixes for contextual KL
Compare 3 commits »
dodox pushed to master at dodox/itpal_jax 2024-12-21 18:31:35 +01:00
de2b9a10d6 Updated README
9fb0014a99 Updated tests (no check kl for w2)
8e991ae05b Fixes
Compare 3 commits »
dodox pushed to master at dodox/fancy_rl 2024-11-07 11:41:18 +01:00
e938018494 trl spec for .reset
4f8fc500b7 Simplify operations on spaces (is_discrete, shape)
5c44448e53 Use trl space definitions (not gym)
8a078fb59e Fix: Issue with env wrapping (ensure batch dim)
52b3f3b71e Updated README
Compare 5 commits »
dodox pushed to master at dodox/fancy_rl 2024-10-21 15:25:30 +02:00
df1ba6fe53 Updated README
8eb9b384c7 Pytest as optional dep
abc8dcbda1 Expand Tests
e927afcc30 Extend policy impl
ca1ee980ef Fix: Bug in loss calc for TRPLLoss
Compare 8 commits »
dodox pushed to master at dodox/NuCon 2024-10-11 08:58:25 +02:00
08ecbb461d Param Repr should contain is_writable not writable
dodox pushed to master at dodox/NuCon 2024-10-10 17:27:23 +02:00
e4c9f047d0 Add PPO example to README
dodox pushed to master at dodox/NuCon 2024-10-10 17:17:40 +02:00
5dfd85a5af Fix README
dodox pushed to master at dodox/NuCon 2024-10-10 17:14:39 +02:00
7e0d85acc7 Update README
878fb9cf4f Fix: Repr for Parameter should contain param_type, not type
81398225ec Ensure we expose the correct default modules
Compare 3 commits »
dodox pushed to master at dodox/NuCon 2024-10-10 16:54:55 +02:00
b750e80c80 Updated README
dodox pushed to master at dodox/NuCon 2024-10-10 16:52:46 +02:00
70fd128465 Expanded README
dodox pushed to master at dodox/NuCon 2024-10-10 16:50:39 +02:00
e625c994df Expanded README
dodox pushed to master at dodox/NuCon 2024-10-10 16:46:22 +02:00
ccbb83674a Fix typo
dodox pushed to master at dodox/NuCon 2024-10-10 16:45:19 +02:00
4791b2a4b6 Updated README
dodox pushed to master at dodox/NuCon 2024-10-10 16:44:05 +02:00
6d1df49ede More sensible bool values for enums
dodox pushed to master at dodox/NuCon 2024-10-10 16:40:20 +02:00
502c8a1c78 Fix typo
dodox pushed to master at dodox/NuCon 2024-10-08 17:21:23 +02:00
66481d8486 Fix typo + better install instructions
dodox pushed to master at dodox/NuCon 2024-10-07 17:11:58 +02:00
c0a9ec33a0 README: Added Note about current capabilities