1. These talks (and the written reports for them) should be directed at an audience of that is fluent in the material you have been studying. The level of the material you are presenting should be somewhat higher than might be found in an introductory level graduate course, (or possibly at the end of one).
2. The style of presentation should be instructional (what you would expect to see in a lecture). The scope of each talk should be an important theorem in the subject with necessary definitions and examples. You should expect to prove the theorem (especially if you haven't had the nerve to do any yet). The written report should be self contained, with an introduction and conclusion (and bibliography), up to twice the length of your other papers. It should be in good technical writing style. (This need not imply dry or boring.)
3. If you have attended a colloquium or a talk at a professional meeting, you should aim for that. If you haven't had the experience, you might ask others for a description of the pacing, which is quite different than what you are used to.
4. In the first few minutes of your presentation and the introduction to your report, relate how this technical talk is an outgrowth or a special aspect of your original choice of seminar topic.
5. The level of this talk should be above that of standard graduate material; the literature sources you use should be expository articles from journals or specialized level books. Most of you are already using sources at this level.
6. Don't forget your announcement should be posted two days before you speak, and set aside time to rehearse your talk.